Skip to main content

Full text of "Proceedings : International Workshop on Subalpine Stone Pines and Their Environment: the Status of Our Knowledge, St. Moritz, Switzerland, September 5-11, 1992"

See other formats


Historic, Archive Document 

Do not assume content reflects current 
scientific knowledge, policies, or practices. 



United^tes 
Department 
Agriculture 



Forest Service 

Intermountain 
Research Station 

General Technical 
Report INT-GTR-309 



July 1994 




Proceedings — International 
Workshop on Subalpine 
Stone Pines and Their 
Environment: the Status 
of Our Knowledge 



n 



:3 



Statements by contributors from outside the U.S. Department 
of Agriculture are solely the views of the authors and do not re- 
flect policies or opinions of the Department. 

The use of trade or firm names in this publication is for reader 
information and does not imply endorsement of any product or 
service by the U.S. Department of Agriculture or other organi- 
zations represented here. 



Intermountain Research Station 
324 25th Street 
Ogden, UT 84401 



Proceedings — International 
Workshop on Subalpine Stone 
Pines and Their Environment: 

the Status of Our Knowledge 

St. Moritz, Switzerland, September 5-11, 1992 



Compilers: 

Wyman C. Schmidt 
Intermountain Research Station 
Forest Service 

U.S. Department of Agriculture 

Friedrich-Karl Holtmeier 
Department of Geography 
University of Munster 
Germany 



Workshop Sponsors: 

intermountain Research Station 
Westfalische Wilhelms-Universitat, 

Landscape Ecology Division, Germany 
Swiss Federal Institute for Forest, Snow, 

and Landscape Research 
Karl and Sophie Binding Foundation, 

Switzerland 




FOREWORD 

This proceedings is a product of the first comprehensive 
and international examination of stone pine species of the 
world. It reports the papers that were presented at the Inter- 
national Workshop on "Subalpine Stone Pines and Their 
Environment — The Status of Our Knowledge" held 
September 5-12, 1992, at St. Moritz, Switzerland. 

It is generally recognized that there are five subalpine 
stone pines in the world— all in the Northern Hemisphere- 
one in North America and four in Europe and Asia. Ail share 
many of the same characteristics and occupy similar ecologi- 
cal niches. Stone pines are defined as those species of Pinus 
that have five-needled fascicles, have wingless seeds, and 
have cones that remain closed at maturity. 

How did this international workshop come about? Comple- 
tion of a whitebark pine workshop in 1987 and a major sym- 
posium in 1989 in Bozeman, MT, U.S.A., sparked interest in 
high-elevation ecosystems of western North America. One 
section of this popular symposium specifically addressed high- 
elevation ecosystems of the world. Information presented 
there hinted that we could learn much about whitebark pine 
ecosystem analogs from other parts of the world that have 
similar physical and biological features. In addition many of 
the Eurasian forests have a long history of human influence. 
The logical place to look for this type of information was the 
world's forest areas that supported subalpine stone pines. 
A cursory examination of the five commonly recognized sub- 
alpine stone pines indicated they had many similarities. 

With this as a background, a small group of western U.S. 
researchers and managers held a planning session with 
Dr. Holtmeier from Muenster, Germany, in Missoula, MT, in 
early 1990. This planning spawned the notion that it was 
time that we started doing a better job of collaborating be- 
tween colleagues in Europe, Asia, and North America. The 
logical starting point was an international workshop that ad- 
dressed those items that held common interest for European, 
Asian, and North American specialists. So, the international 
planning transmission was put into gear and 2V2 years later a 
workshop was held in St. Moritz, Switzerland. 

The primary objectives of the Stone Pine Workshop were: 

1 . To improve international cooperation and collaboration 
between scientists. 

2. To exchange research findings. 

3. To determine knowledge gaps in the ecology and man- 
agement of subalpine stone pines of the world. 

All five subalpine stone pines were covered by authors of 
48 papers from Europe, Asia, and North America. They were: 



Species 

Swiss stone pine 
{Pinus cembra) 

Siberian stone pine 
{Pinus sibirica) 



Natural range 

European Alps, Carpathian 
Mountains 



Siberian Russia and Northern 
Mongolia (95 percent in 
Russia, 5 percent in Mongolia) 

Japanese stone pine Japan, Korea, Eastern Siberia 
{Pinus pumila) 

Korean stone pine Northeastern China, North Korea, 
{Pinus koraiensis) Honshu Japan, southeastern 

Russia 



Whitebark pine 
{Pinus albicaulis) 



Western North America 



The 1 2 countries represented and the number of papers 
from each country were: Russia (9), China (1), Japan 
(4), Switzerland (5), Austria (4), Italy (3), France (1), Yugosla- 
via (1), Czechoslovakia (1), Romania (1), Germany (3), and 
United States (15). The main subjects covered were: 

• Evolution and taxonomy. 

• Basic ecology of stone pine species. 

• Growth characteristics. 

• Influence of environmental factors. 

• Regeneration. 

• Importance to wildlife. 

• Forest structure and dynamics. 

• Forest management. 

• Research needs. 

Stone pines are critical forest components on vast areas. 
Pinus sibirica alone covers 40 million hectares, mainly in 
Siberia, and is a key component of the economy and biology. 
P. l<oraiensis accounts for one-fourth of the timber production 
in China. P. albicaulis is key to the survival of wildlife such as 
the grizzly bear and is a protector of high-elevation water- 
sheds and visual resources. P. cembra plays a key role in 
avalanche protection, wood production, and recreation in 
Europe. P. pumila is a key high-elevation species for stabili- 
zation of mountain slopes in Japan and Kamchatka. These 
species have many things in common, likely starting with a 
common ancestor, probably P. sibirica. They also share 
common methods of seed dispersal with the North American 
and Eurasian nutcrackers. Natural selection processes have 
resulted in a real mutualism between subalpine stone pines 
and the nutcrackers. 

This workshop was sponsored by the Intermountain 
Research Station, USDA Forest Service, U.S.A.; the 
Westfalische Wilhelms-Universitat, Landscape Ecology Divi- 
sion, Germany; the Swiss Federal Institute for Forest, Snow, 
and Landscape Research, Switzerland; and the Karl and 
Sophie Binding Foundation, Switzerland. 

We want to acknowledge the people and the organizations 
that supported them for their enthusiastic and dedicated ef- 
forts in the planning and conduct of this international work- 
shop. Three field tours were conducted by the hosts of the 
workshop. The first tour entitled "Engadine Cembra Pine- 
Larch Forest," was in a cembra pine-larch forest overlooking 
St. Moritz, a heavily used recreation area in the Engadine. 
The primary subjects were tree regeneration and succession 
in avalanche tracks and, at the upper timberline, squirrel and 
nutcracker interactions, human historical influence, grazing 
and game, and the pressures of modern tourism on the for- 
ests and landscape. 

The second tour, entitled "Ecology and Technique of High- 
Altitude Afforestation," dealt primarily with management of 
avalanche-prone areas. It was conducted in the Stillberg Ex- 
perimental Area near timberline in the Dischma Valley near 
Davos. It demonstrated the use and design of physical barri- 
ers for avalanche control and reforestation methods with 
cembra pine, larch, and mugo pine for long-term reduction of 
avalanche problems. 

The third tour, entitled "Subalpine and Alpine Environ- 
ments," emphasized vegetation succession, particularly of 
cembra pine and larch following the gradual retreat of the 
Morteratsch Glacier. Also included in this tour was a visit to 
the glacier areas at Diavolezza on the Bernina Pass where 
recreation values are paramount. From the aerial tram to the 



mountain station (2,973 m) we viewed the strong contrasts of 
vegetation and glaciers on the north slopes of the Bernina 
Pass area. 

For all the hosts of the enlightening field tours and arrange- 
ments at St. fvloritz we owe a special thanks. Those included 
Drs. W. Schonenberger, R. Hasler, J. Senn, R. Lassig, 
F.-K. Holtmeier, and Herr U. Wasem — also Engadine 
Foresters J. Altman, I. A. Bisaz, and G. Bott. We also thank 
Prof. Dr. H. Steinlin from the Binding-Siftung who attended 
the workshop as representative of the Binding Foundation. 
All of these people and their staffs were wonderful hosts who 
contributed so much to the success of this first international 
conference on stone pines. 

A special thanks goes to Kathy McDonald, Intermountain 
Research Station, U.S.A., who processed all the manuscripts. 

The organizing committee was composed of the following 
people: 

Dr. W. Schonenberger 
Dr. R. Hasler 
Dr. J. Senn 
Swiss Federal Institute for 
Forest, Snow and Landscape Research 
Zurcher Str. 1 1 1 
CH-8903 Birmensdorf 
Switzerland 

Prof. Dr. F.-K. Holtmeier 

University of f^unster 
Department of Geography 
Landscape Ecology Division 
Robert-Koch-Str. 26 
W-4400 Munster 
Germany 

Dr. Wyman C. Schmidt 
USDA Forest Service 
Intermountain Research Station 
Forestry Sciences Laboratory 
Montana State University 
Bozeman, MT, 59717-0278, U.S.A. 



Drs. Holtmeier and Schmidt served as technical evaluators 
of all the papers. 

As a consequence of this international workshop, a resolu- 
tion was passed by the scientists in attendance that aimed to- 
ward greater coordination and sharing of research findings. 
The resolution was: 

1 . To organize an International Association for Stone Pine 
Research. 

2. To form an international organizing committee by elec- 
tion that could probably meet in Tomsk, Russia, in 1993 (or 
later) to (a) establish the objectives of the Association; and 
(b) establish the organizational structure of the Association. 

3. The organizing committee to develop the preliminary 
plan for the next International Stone Pine Workshop. 

4. To hold the next International Stone Pine Workshop at 
Tomsk, Russia, in 1995 (or later). 

The world faces many environmental and resource chal- 
lenges. Many of these challenges are in temperate forests of 
the Northern Hemisphere. In these, stone pines are a signifi- 
cant forest component and one in which our knowledge base 
is certainly not complete. Sharing the knowledge developed 
from similar ecosystems and building on that solid base is 
key to a better understanding of how our subalpine forests 
function ecologically. We can learn much from each other. 
The international workshop that generated this proceedings is 
a good first step in that direction. Let us do what we can to- 
gether to keep the momentum. 



Wyman C, Schmidt F.-Karl Holtmeier 



VORWORT 



Bei den Arten handelte es sich um: 



Dieser Band ist eine erste umfassende internationale Studie 
iiber die subalpinen Steinkiefern der Erde. Er enthalt die 
Vortrage, die anipiich des Internationalen Workshops "Subal- 
plne Stone Pines and Their Environment — The Status of Our 
Knowledge" gehalten wurden. Dieser Workshop fand vom 
5. bis 1 1 . Septennber 1 992 in St. Moritz (Schweiz) statt. Die 
funf subalpinen Steinkieferarten sind ausschliepiich in der 
Nordhemisphare verbreitet — davon eine in Nordamerika 
und vier in Eurasien. Diese Arten weisen viele gemeinsame 
Merkmale auf und besetzen auch uhnliche okologische 
Nischen. Unter Steinkiefern versteht man die Arten des 
Genus Pinus mit funfnadeligen Kurztrieben, flugellosen 
Samen und Zapfen, die im Reifezustand geschlossen 
bleiben. 

Die Vorgeschichte dieses Internationalen Workshops 
beginnt 1 987. Damals fand in Bozeman ein Workshop uber 
die Whitebark Pine {Pinus albicaulis) statt, dem 1989 ein 
gropes Symposium, ebenfalls in Bozeman, folgte. Beide 
Veranstaltungen stiepen auf eine grope Resonanz und fuhrten 
zu einem wachsenden Interesse an den Hochgebirgsoko- 
systemen des westlichen Nordamerika. Eine Vortragssektion 
dieses Symposiums war den Hochgebirgsokosystemen der 
Welt gewidmet. Dabei zeigte sich, dap aus einem Vergleich 
mit Steinkiefer-Okosystemen anderer in ihrer Naturaus- 
stattung mit dem nordamerikanischen Verbreitungsareal der 
Whitebark Pine vergleichbarer Regionen der Erde auch 
wichtige Informationen zum besseren Verstandnis der 
Whitebark Pine - Walder gewonnen werden konnen. Als ein 
wesentlicher Unterschied zu den amerikanischen Whitebark 
Pine - Waldern erwies sich die zum Tell schon einige tausend 
Jahren zuruckreichende anthropogene Beeinflussung der 
eurasiatischen Steinkiefernwalder durch den Menschen. Mit 
Sicherheit lassen sich aus den dortigen Verhaltnissen auch 
nutzliche Hinweise auch potentielle Auswirkungen men- 
schlicher Eingriffe in den amerikanischen Whitebark Pine - 
Waldern gewinnen. 

Anfang 1990 traf sich dann eine kleine Gruppe von Wissen- 
schaftlern und Forstleuten aus den westlichen Vereinigten 
Staaten mit Dr. Holtmeier (Munster, Deutschland) in Missoula 
(Montana), um uber weiterfuhrende Forschungen iiber die 
subalpinen Steinkiefern zu beraten. Der nachste Schritt 
sollte eine Verbesserung und Intensivierung der wissen- 
schaftlichen Zusammenarbeit der Fachleute aus Europa, 
Asien und Nordamerika sein. Man kam uberein, ein en 
Internationalen Workshop zu veranstalten, mit dem Ziel, dort 
Fragen von gemeinsamem Interesse zu diskutieren und neue 
Forschungsansatze zu entwickeln. Unmittelbar nach diesem 
Treffen begannen die Vorbereitungen zu diesem Workshop, 
der dann zweieinhalb Jahre spater in St. Moritz (Schweiz) 
stattfand. 

Seine Hauptziele waren: 

1 . Verbesserung und Intensivierung der internationalen 
Zusammenarbeit. 

2. Austausch von Forschungsergebnissen. 

3. Forschungsbedarf hinsichtlich der Okologie und 
Bewirtschaftung subalpiner Steinkiefernwalder aufzuzeigen. 

Uber alle funf Steinkieferarten wurden Vortrage gehalten. 
Insgesamt waren es 48 Referate, an denen Wissenschaftler 
aus 12 Staaten beteiligt waren. 



Art 

Arve (Synonym: Zirbe) 
{Pinus cembra) 

Sibirische Arve 
{Pinus sibirica) 



Sibirische Zwergarve 
(Japanische 
Steinkiefer) 
{Pinus pumila) 

Koreanische Steinkiefer 
{Pinus lioraiensis) 

Weiprinden-Kiefer 
{Pinus albicaulis) 



Naturliches 
Verbreitunggebiet 

Europaische Alpen, Karpaten 



Russisch Sibirien und ndrdliche 
Mongolei (95 percent in 
Rupiand; 5 percent in der 
Mongolei) 

Japan, Korea, Ostsibirien 



Nordost-China, Nordkorea, 
Honshu, Siidostasien 

Westliches Nordamerika 



Folgende Lander waren vertreten (Zahl der Beitrage in 
Klammern): Rupiand (9), China (1), Japan (4), Schweiz (5), 
Osterreich (4), Italien (3), Frankreich (1), Yugoslawien (1), 
Tschechoslowakei (1), Rumanien (1), Deutschland (3) und 
USA (15) 

Die Hauptthemen waren: 

• Evolution und Taxonomie. 

• Autokologie der subalpinen Steinkieferarten. 

• Wachstumscharakteristik. 

• Einflup der Standortfaktoren. 

• Regeneration. 

• Bedeutung der Steinkiefern fiir die Tiere. 

• Waldstrukturen und -dynamik. 

• Waldbewirtschaftung und -pflege. 

• Forschungsbedarf. 

Die Steinkiefern stellen einen hohen Anteil an der Wald- 
bedeckung groper Gebiete. Allein Pinus sibirica bedeckt ein 
Areal von 40 Millionen Hektar. Der gropte Teil davon liegt in 
Sibirien, wo der Sibirischen Arve sowohl in ihrer Bedeutung 
fur die Wirtschaft als auch im Hinblick auf die biologische 
Situation die Rolle eines Schlusselfaktors zukommt. Pinus 
koraiensis stellt ein Viertel der Holzproduktion in China. 
Pinus albicaulis ist unter anderem fur den Grizzly-Bar von 
existentieller Bedeutung. Auperdem erfullt diese Art in hoch- 
gelegenen Wassereinzugsgebieten eine wichtige Schutz- 
funktion und erhoht zudem wesentlich den asthetischen Reiz 
der Gebirgslandschaft. Pinus cembra kommt in Europa eine 
wichtige Rolle bei der Lawinenvorbeugung durch Schutz- 
walder zu. Zugleich liefert sie ein fiir Tafelungen und Schnitz- 
erein geschatztes Holz. Auch sie ist von hohem asthetischen 
Wert. Pinus pumila tragt in Japan und Kamtschatka ent- 
scheidend zur Stabilisierung steiler Gebirgshange bei. Die 
vielen Gemeinsamkeiten der genannten Arten haben ihre 
Ursache vermutlich in ihrer Abstammung von einer Art, 
moglichenweise von Pinus sibirica. Charakteristisch ist fiir sie 
auch die Verbreitung ihrer Samen durch den nordamerika- 
nischen und europaischen Tannenhaher. Zwischen den 
subalpinen Steinkiefern und den Hahern hat sich im Laufe 
der Evolution ein echter Mutualismus entwickelt. 



Der Workshop wurde unterstutzt durch die Intermountain 
Research Station, USDA Forest Service, USA, die West- 
falische Wilhelms-Universitat (MOnster, Deutschland), die 
Eidgenossische Forschungsanstait fur Wald, Schnee und 
Landschaft (Schweiz) sowie durch die Karl und Sophie Bind- 
ing - Stiftung. 

Wir danken alien Personen und Organisationen, die mit 
ihrem Enthusiasmus und ihrer tatkraftigen Unterutzung die 
Planung und die Durchfuhrung dieses Workshops ermbglicht 
haben. 

Wahrend des Workshops fanden unter der Leitung der 
Gastgeber drei Exkursionen statt. Die erste fuhrte in die 
Larchen-Arvenwalder oberhalb des bekannten Engadiner 
Hohenkurortes St. Moritz. Hauptthemen waren die Ver- 
jungung und Sukzession in Lawinenzugen und im Wald- 
grenzbereich, die Einbindung von Eichhornchen und Tannen- 
haher in das Okosystem des Larchen-Arvenwaldes, die 
Einflusse von Beweidung und Wild sowie die Belastungen 
der Walder und der Landschaft durch den modernen 
Tourismus. 

Die zweite Exkursion, die sich im wesentlichen mit okologi- 
schen Problemen und Techniken bei der Hochlagenauf- 
forstung und der Behandlung von Lawineneinzugsgebieten 
befapte, fuhrte in die im Waldgrenzbereich gelegenen Ver- 
suchsflachen am Stillberg im Dischmatal bei Davos. Dort 
wurden kunstliche Lawinenverbauungen und Wieder- 
aufforstungen mit Arven, Larche und Legfohre als Mittel zur 
Lawinenvorbeugung demonstriert. 

Die dritte Tour, die unter dem Thema "subalpine und alpine 
Lebensraume" stand, fuhrte zunachst in das Morteratsch-Tal. 
Dort ging es vor allem um die Vegetationssukzession im 
Vorfeld des allmahlich zuruckweichenden Morteratsch- 
Gletschers unter besonderer Berucksichtigung der 
Besiedlung durch Arve und Larche. Mit einem Besuch der 
Diavolezza am Berninapap fand die Exkursion ihren 
Abschlup. Von der Bergstation (2973 m) der Luftseilbahn bot 
sich ein uberwaltigender Blick auf die stark vergletscherte 
Nordseite der Berninagruppe. 

Fur alles, die eriebnisreichen und lehrrreichen Exkur- 
sionen sowie die Veranstaltungen in St. Moritz, schulden wir 
den Gastgebern — Drs. W. Schonenberger, R. Hasler, 
J. Senn, R. Lassig, F.-K. Holtmeier und Herr U. Wasem — 
sowie den Engadiner Fbrstern J. Altmann, I. A. Bisaz, und 
G. Bott besonderen Dank. Nicht zuletzt geht unser Dank 
auch an Herrn Prof. Dr. H. Steinlin, der als Vertreter der 
Binding-Stiftung selbst an dem Workshop teilnahm. Sie alle 
haben mit ihrem gro^en Engagement bei der Planung und 
der Durchfuhrung zum Gelingen dieses ersten internationalen 
Workshops uber die Steinkiefern beigetragen. 

Besonderer Dank gilt auch Frau Kathy McDonald und 
Herrn Richard Klade (Intermountain Research Station, USA) 
fur die redaktionelle Bearbeitung der Manuskripte. 

Das Organistaionskomitee setzte sich zusammen aus: 

Dr. W. Schonenberger 
Dr. W. Hasler 
Dr. J. Senn 
Eidgenossische Forschungsanstait 
fur Wald, Schnee und Landschaft 
Zurcher Strape 1 1 1 
Ch-8093 Birmensdorf 
Schweiz 



Prof. Dr. F.-K. Holtmeier 
Westfalische Wilhelms-Universitat 
Institut fur Geographie 
Abteilung Landschaftsokologie 
Robert-Koch-Strape 29 
D-48149 Munster 
Deutschland 

Dr. Wyman C. Schmidt 
USDA Forest Service 
Intermountain Research Station 
Forestry Sciences Laboratory 
Montana State University 
Bozeman, Montana, USA 59717-0278 

Die Vorbereitung der Manuskripte fur die redaktionelle 
Bearbeitung wurde von Dr. Holtmeier und Dr. Schmidt 
durchgefuhrt. 

Wahrend des Workshops veranschiedeten die Teilnehmer 
eine Resolution, die eine noch weitergehende Koordination 
der Forschung und bessere gegenseitige Information uber 
neue Forschungsergebnisse zum Ziel hat. Sie umfapt 
folgende Punkte: 

1 . Grundung einer Internationalen Vereinigung fur die 
Erforschung der Steinkiefern. 

2. Bildung eines Internationalen Organisationskomitees, 
das erstmals 1993 (oder spater) in Tomsk (Rupiand) 
zusammentreten soli, um die kunftigen Ziele dieser Ver- 
einigung zu formulieren und eine Organisationstruktur zu 
schaffen. 

3. Planung des nachsten Internationalen Workshops uber 
die Steinkiefern. 

4. Durchfuhrung dieses Workshops voraussichtlich 1995 
(Oder spater) in Tomsk. 

Die Welt sieht sich vielen Herausforderungen durch 
Umweltprobleme und Ressourcennutzung gegenuber. 
Davon sind auch die Gebirgswalder und borealen Walder der 
Nordhemisphare betroffen, zu denen auch die Steinkiefern- 
walder gehoren. Das Wissen uber sie ist noch luckenhaft. 
Indem wir unser Wissen zusammentragen bilden wir eine 
solide Grundlage, von der aus es leichter sein wird, tiefere 
Einsichten in die okologischen Wirkungsgefuge dieser 
Waldokosysteme zu gewinnen. Der Internationale Workshop 
in St. Moritz mit den hier zusammengestellten Beitragen ist 
ein erster Schritt in diese Richtung. Jetzt geht es darum, 
diesen Weg weiterzugehen. 



Wyman C. Schmidt F.-Kari Holtmeier 



CONTENTS 



AUTHORS TITLE PAGE 

Wyman C. Schmidt Distribution of Stone Pines 1 

FIELD TOURS 

Swiss Hosts Tour Descriptions 8 

Mihailo Grbic 

Friedrich-Karl Holtmeier Introduction to the Upper Engadine and Its Forest 9 

EVOLUTION AND TAXONOMY 
Konstantin V. Krutovskli Genetic Differentiation and Phylogeny of Stone Pine Species Based on 

Dmitri V. Politov Isozyme Loci 19 

Yuri P. Altukhov 

Hermann Mattes Coevolutional Aspects of Stone Pines and Nutcrackers 31 

Dmitri V. Politov Allozyme Polymorphism, Heterozygosity, and Mating System of Stone Pines 36 

Konstantin V. Krutovskii 

Diana F. Tomback Genetic Population Structure and Growth Form Distribution in Bird-Dispersed Pines 43 

William S. F. Schuster 

BASIC ECOLOGY 

Ernst Frehner Experiences With Reproduction of Cembra Pine 52 

Walter Schonenberger 

P. L. Gorchakovsky Distribution and Ecology of Siberian Stone Pine in the Urals 56 

Rudolf Hasler Ecophysiological Investigations on Cembran Pine at Timberline in the Alps, An Overview 61 

Peter A. Khomentovsky A Pattern of Pinus pumila Seed Production Ecology in the Mountains of Central 

Kamtchatka 67 

Takayoshi Koike Needle Longevity and Photosynthetic Performance in Cembran Pine and Norway Spruce 

Rudolf Hasler Growing on the North- and East-Facing Slopes at the Timberline of Stillberg in the 

Hans Item Swiss Alps 78 

Luo Ju Chun The Broad-leaved Korean Pine Forest in China 81 

Tad Weaver Climates Where Stone Pines Grow, A Comparison 85 

GROWTH CHARACTERISTICS 

Kurt Holzer Growth of Swiss Stone Pines That Originated From and Were Planted at Several 

Altitudes in the Austrian Alps 91 

Takuya Kajimoto Seasonal Patterns of Growth and Photosynthetic Activity of Pinus pumila Growing on the 

Kiso Mountain Range, Central Japan 93 

Herbert Kronfuss Height Growth in Cembran Pine as a Factor of Air Temperature 99 

Josef Senn Survival and Growth of Planted Cembran Pines at the Alpine Timberline 105 

Walter Schbnenberg 
Ueli Wasem 



AUTHORS TITLE PAGE 

INFLUENCE OF ENVIRONMENTAL FACTORS 

Mihailo Grbic Changes of Swiss Stone Pine Aphid Life Cycle, Density, and Population Structure in 

High-Altitude Swiss Stone Pine Afforestation 111 

Raymond J. Hoff Genetic Consequences and Research Challenges of Blister Rust in Whitebark Pine 

Susan K. Hagle Forests 118 

Richard G. Krebill 

Karel Kahik Performance of Pinus cembra, P. peace, and P. strobiformis Within Air-Polluted Areas 127 

Todd Kipfer Competition and Crown Characteristics of Whitebark Pine Following Logging in 

Katherine Hansen Montana, U.S.A 130 

Ward McCaughey 

Penelope Morgan Fire Ecology of Whitebark Pine Forests of the Northern Rocky Mountains, U.S.A 136 

Stephen C. Bunting 
Robert E. Keane 
Stephen F. Arno 

Tad Weaver Vegetation Distribution and Production in Rocky Mountain Climates — With Emphasis 

on Whitebark Pine 142 

REGENERATION 

I. Blada Variation in Size and Weight of Cones and Seeds in Four Natural Populations of 

N. Popescu Carpathian Stone Pine 154 

Mitsuhiro Hayashida Role of Nutcrackers on Seed Dispersal and Establishment of Pinus pumila and 

P. pentaphylla 1 59 

Harry E. Hutchins Role of Various Animals in Dispersal and Establishment of Whitebark Pine in the 

Rocky Mountains, U.S.A 163 

Hermann Mattes Size of Pine Areas in Relation to Seed Dispersal 172 

Ward W. McCaughey The Regeneration Process of Whitebark Pine 179 

Ikuko Nakashinden Japanese Stone Pine Cone Production Estimated From Cone Scars, Mount 

Kisokomagatake, Central Japanese Alps 188 

Diana F. Tomback Effects of Seed Dispersal by Clark's Nutcracker on Early Postfire Regeneration of 

Whitebark Pine 193 

Vladislav N. Vorobjev New Trend in Dendrochronology: 1. Theoretical Principles of Reprochronology 199 

V. N. Vorobjev New Trend in Dendrochronology: 2. Method of Retrospective Study of Seminiference 

S. N. Goroshkevich Dynamics in Pinaceae 201 

D. A. Savchuk 

IMPORTANCE TO WILDLIFE 

Ronald M. Lanner Nutritive Value of Whitebark Pine Seeds, and the Question of Their Variable Dormancy 206 

Barrie K. Gilbert 

David J. Mattson Bear Use of Whitebark Pine Seeds in North America 212 

Daniel P. Reinhart 

Diana F. Tomback Ecological Relationship Between Clark's Nutcracker and Four Wingless-Seed Strobus 

Pines of Western North America 221 



AUTHORS TITLE PAGE 

FOREST STRUCTURE AND DYNAMICS 

Giovanna De Mas Structure of Swiss Stone Pine Stands in Northeastern Italy 226 

Elena Piutti 

S. Yu. Grishin Role of Pinus pumila in Primary Succession on the Lava Flows of Volcanoes of 

Kamchatka 240 

Robert E. Keane Decline of Whitebark Pine in the Bob Marshall Wilderness Complex of Montana, U.S.A 245 

Penelope Morgan 

Renzo Motta Some Aspects of Cembran Pine Regeneration in the Italian Cottian Alps 254 

Alberto Dotta 

Pietro Piussi Mixed Cembran Pine Stands on the Southern Slope of the Eastern Alps 261 

Siegfried Sauermoser Current Distribution of Cembra Pine in the Lechtal Alps 269 

E. P. Smolonogov Geographical Differentiation and Dynamics of Siberian Stone Pine Forests in Eurasia 275 

Use WI6rick The Cembran Pine in the French Alps: Stand Dynamics of a Cembran Pine Forest 

in Tueda (Savoy, France) 280 

FOREST MANAGEMENT 

Werner Frey Silvicultural Treatment and Avalanche Protection of Swiss Stone Pine Forests 290 

Riet Gordon Importance and Silvicultural Treatment of Stone Pine in the Upper Engadine (Grisons) 294 

Jorg Heumader Cultivation of Cembran Pine Plants for High-Elevation Afforestations 298 

Katherine C. Kendall Whitebark Pine Conservation in North American National Parks 302 

RESEARCH NEEDS 

Wyman C. Schmidt Research Needs in Whitebark Pine Ecosystems 309 

Vladislav N.Vorobjev Problems of Comprehensive Investigation, Utilization, and Reproduction of 

Nina A. Vorobjeva Russian Cedar Pine Forests 314 



DISTRIBUTION OF STONE PINES 

Wyman C. Schmidt 



The five stone pine species that are the subject of this 
proceedings occur only in the Northern Hemisphere — one 
species in North America and the other four in Europe 
and Asia. Taxonomically they fall in section Strobus and 
subsection Cembrae. They are characterized by having 
five-needled fascicles, cones that do not open at maturity, 
and seeds that are wingless. 

The distributions of the five species are reasonably well 
defined, but updates on their distribution usually show 
minor changes. Also, in some cases, such as with white- 
bark pine (Pinus albicaulis), the species occurs only at 
high-elevation sites, and there may be long distances be- 
tween the mountain ranges that support this species. This 
species ocoirs over a large geographic area; however, there 
are usually large intervening valleys and lowlands not 
suitable for whitebark pine forests. The same is true for 
the other stone pines. Therefore, just because a distribu- 
tion boundary of a species encompasses a large area, it 
does not mean that all the mountains and valleys are 
fully occupied by that species. 

For the purposes of this proceedings, we have chosen to 
show the general distribution on the maps of the five spe- 
cies and do not necessarily include all of the individual 
outliers. 

The distribution maps that follow are adapted fi'om those 
pubUshed by Critchfield and Little (1966). The maps show 
that the five species are mostly discrete populations, but 
there are overlapping distributions of Siberian stone pine 
(Pinus sibirica) and Japanese stone pine (P. pumila), and 
an overlap of Japanese stone pine and Korean stone pine 
(P. koraiensis). 

Whitebark pine occurs only in North America at high 
elevations fi*om central British Columbia south through 
the Cascade Mountains of Washington and Oregon to the 
southern Sierra Nevada of California, in a limited area of 
V Nevada, and extensively in Montana, Idaho, and western 
Wyoming. It occurs just below the timberline and extends 
downward elevationally into associations with several 
other conifers. 

Swiss stone pine grows at high elevations in the Alps and 
Carpathian Moimtains. It occurs mostly in Switzerland, 
Austria, northern Italy, southeastern France, Romania, 
Czechoslovakia, and Poland. Although Swiss stone pine 
is considered closely related to Siberian stone pine, over 
2,000 kilometers presently separate the two species. 



Swiss stone pine commonly grows in association with 
other conifers. 

Siberian stone pine is widely distributed in Eurasia fi"om 
the Ural Moimtains through western and central Siberia to 
northern MongoUa — an east-west distance of about 4,000 
kilometers. It overlaps the range of Japanese stone pine 
on the eastern portion of its distribution near Lake Baikal 
in Siberia. Different than both whitebark pine and Swiss 
stone pine, which only occupy high-elevation sites, Sibe- 
rian stone pine (usually called cedar-pine or Siberian ce- 
dar by Russians) commonly grows in the plains and river 
valleys, as well as in the mountain areas. 

Japanese stone pine grows in northeastern Siberia and 
in subalpine parts of Japan, Korea, and Manchvuia. It ex- 
tends north almost to the Arctic Ocean (70° N.), west to 
Mongolia and to the Lake Baikal area, and south to Korea 
and Honshu, Japan. It grows from low elevations in the 
north to high elevations in the south. It commonly forms 
thickets just below barren tundra on tops of mountains or 
lowlands in the extreme north. Like Siberian stone pine, 
it extends nearly 4,000 kilometers fi*om east to west and 
3,000 kilometers north to south. 

Korean stone pine occurs throughout Korea and eastern 
Manchuria into southeastern Siberia as well as on the is- 
lands of Honshu and Shikoku in Japan. It tends to grow 
in more maritime conditions and lower moimtains than its 
neighbor — the Japanese stone pine. Its geographic range 
overlaps that of the Japanese stone pine, but the species 
occupy different ecological niches. Korean stone pine, im- 
like its counterparts, will grow in association with hard- 
woods where near-maritime cUmates prevail. Its valuable 
wood makes it a much sought after species. 

The following maps illustrate the geographic distribu- 
tion of the five stone pines. These descriptions are based 
primarily on those of Critchfield and Little (1966) and 
Mirov (1967). 

References 

Critchfield, W. B.; Little, E. L., Jr. 1966. Geographic dis- 
tribution of the pines of the world. Misc. Publ. 991. 
Washington, DC: U.S. Department of Agricvdture, For- 
est Service. 97 p. 

Mirov, N. T. 1967. The genus Pinus. New York: Ronald 
Press Company. 602 p. 



1 



NORTH AMERICA 



135° 130° 125° 120° 115° 110° 




130° 125° 120° 115° 110° 105° 100° 



2 



EUROPE 



10° 5° 0° 5° 10° 15° 20° 25° 30° 35° 




0° 5° 10° 15° 20° 25° 30° 35° 



3 



ASIA 




4 



ASIA 




5 



ASIA 




Field Tours 




International Workshop 
St. Moritz 1 992 



Field tours were an important component of the Stone Pine Workshop. First, 
the tours provided an informal setting for scientists from the 12 countries to become 
acquainted with each other and their research; second, the tours demonstrated the 
research and management of Swiss stone pine forests in the rugged and spectacularly 
beautiful mountains of Switzerland. 

The following section is the handout from the Swiss hosts used to describe 
the three major field tours. The drawings by a gifted scientist, Mihailo Grbic, from 
Belgrade, Yugoslavia, depict a symbolic participant, Dr. Stone Pine, enjoying the rather 
strenuous excursions. Grbic also created the welcome drawing used as sections 
division indicators in this publication. 



7 



TOUR DESCRIPTIONS 

Swiss Hosts 



We will try to use pleasant weather for the field tours. 
Therefore this preliminary schedule may change a little. 
Usually weather is fine in September. However, snow and 
fi-ost are possible. Do not forget warm clothing, tough foot- 
wear, backpack, sunblock, and simglasses. During the field 
toiu-s we will have lunch fi-om the backpack. 

Engadine Cembra Pine-Larch Forest 

Sunday 6th. St. Moritz, Muottas da Schlarigna. Walk 
fi-om St. Moritz Bad (Hotel "Laudinella'')-Stazer Wald- 
Muottas da Schlarigna-St. Moritz Bad (1,800 to 2,300 m 
above sea level [a.s.l.]). 

Cembra pine-larch forest: composition, structure, growth 
characteristics, old cembra pines, regeneration pattern, 
understory vegetation, avalanche tracks with green alder 
{Alnus viridis) formation, influence of seed-caching ani- 
mals (squirrel, nutcracker), effects of seed dispersal by 
the thick-billed European nutcracker (Nucifraga c. 
caryocatactes). 

Human influence on the forest distribution pattern, 
human-caused timberline, influence of grazing livestock 
and game, and of toiu"ism; invasion of rarely used and 
abandoned alpine and forest pastures by trees, influence 
of microtopography and microclimate on alpine vegetation 
and on growth of invading trees. 

Ecology and Technique of High- 
Altitude Afforestation 

Tuesday 8th. Davos, Stillberg. Bus fi"om St. Moritz- 
Albula pass-Davos-Teufi (Dischma Valley, 1,700 m), 
5-hour walk to and through Stillberg experimental area 
near timberline at 2,000 to 2,230 m a.s.l. and back to the 
valley bottom (Teufi). 

The experimental area lies on a steep avalanche catch- 
ment slope. Construction of avalanche barriers. Dam- 
ages, diseases, mortality and development of a 17-year-old 




plantation of Pinus cembra, Larix decidua, and P. mugo. 
Influence of different site factors such as wind, tempera- 
ture, irradiation, snow movement, duration of snow cover, 
soil properties, vegetation on plantation success. Eco- 
physiology of subalpine timberline trees. Conclusions for 
plantation technique and spatial patterns. 




Subalpine and Alpine Environments 

Friday 11th. Morteratsch Glacier, Diavolezza. 
Morteratsch Glacier (2,000 m, 2- to 3-hour walk). Train to 
Morteratsch station, walk to Morteratsch Glacier via Alp 
Chiinetta (2,052 m); old-growth cembra pine forests, re- 
generation, invasion of the lateral moraine (fi-om 1850) 
and the glacier forefield by cembra pine and larch. 

Diavolezza (2,973 m). Train fi-om Morteratsch station 
through the Bemina Valley to Diavolezza station (2,200 m) 
on Bemina Pass, and by ropeway to Diavolezza. Magnifi- 
cent view of the Bernina Massive (4,049 m) with Pers Gla- 
cier and Morteratsch Glacier. 

★ ***★ 

In addition to providing valuable scientific and manage- 
ment information, the introductory paper by Friedrich- 
Karl Holtmeier that follows gave background and baseline 
data describing the local Swiss stone pine forests that 
were the focus of the field tours. 

***** 



8 



INTRODUCTION TO THE UPPER 
ENGADINE AND ITS FOREST 

Friedrich-Karl Holtmeier 



Abstract — The Engadine is characterised by great mean eleva- 
tion (2,400 m) and a relatively continental climate. Cembra pine 
{Pinus cembra) and European larch (Larix decidua) are the most 
important tree species. The area was already settled in prehis- 
toric time. Since then, mountain forest has been influenced by 
human impact. During history timberline became lower by about 
150-300 m. At present, the upper forest limit is located at about 
2,200-2,300 m. Composition and structure of the forest also 
changed in that larch could spread at the cost of cembra pine on 
all areas that were suitable for grazing. After grazing pressure 
has declined natural succession is revived, and, in the long term, 
larch will be replaced by cembra pine, if not specially managed. 
Natural stands of Norway spruce (Picea ahies) occur on the 
northwest-exposed slopes of the main valleys within an area 
characterized by higher humidity. Spruce reaches its upper limit 
at 2,000 m. Dwarfed mountain pine (Pinus mugo) and green al- 
der (Alnus viridis) are most common in avalanche tracks. Green 
alder is to be found mainly on moist sites, whereas mountain 
pine prevails under dry conditions. Alder also occurs in bogs. 



The Engadine — Engiadina in Rhacto-Romsinic lan- 
guages — is located in the eastern part of Switzerland and 
comprises the uppermost drainage area of the Inn River 
(fig. 1). The main valley trends from southwest to north- 
east. By an administrative border the Engadine is divided 
into Upper and Lower Engadine. However, landscape 
physiognomy of both sections is also different. The Lower 
Engadine is deeply cut into the mountains, and mountain 
slopes are steep and nagged. The Upper Engadine main 
valley, on the other hand, is wide and topography is com- 
paratively smooth. Between St. Moritz and the Maloja 
Pass the valley is characterized by four lakes separated by 
alluvial plains. To the southeast the Bemina Valley gives 
access to the famous glaciated Bemina Group. 

The Engadine is characterized by great mean elevation 
(2,400 m). The bottom of the main valley is located at an 
altitude of about 1,800 m. The tributary valleys climb up 
to about 2,300 m. The highest peaks' elevations are about 
4,000 m (Piz Bernina 4,049 m; fig. 1). In view of its great 
elevation the Upper Engadine has been called The Roof 
of Europe" by the English. 

Due to the geographical location in the central Alps 
and the high mass-elevation of the Engadine the climate 
is rather continental, characterized by a relatively great 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Friedrich-Karl Holtmeier is Professor of Landscape Ecology and Geogra- 
phy, Department of Geography, University of Munster, Germany. 



amplitude of mean temperature and comparatively low 
amount of precipitation (fig. 2). Annual precipitation de- 
creases from about 1,400 mm at Maloja Pass, where the 
main valley is open to moisture-carrying air ciurents from 
the southwest (fig. 3), to less than 1,000 mm at St. Moritz. 
At timberline, however, precipitation is much higher than 
at the valley bottom, as can be concluded from the data 
available for Julier Hospiz on Julier Pass and for Bemina 
Hospiz on Bemina Pass. At both stations, anniial precipi- 
tation exceeds 1,500 mm. 

Due to the continental climate, the upper limits of veg- 
etation and cultivation reach relatively high altitudes. In 
the Upper Engadine, forest limit is located at about 2,200- 
2,300 m. Solitary crippled trees may still be found at and 
even above 2,500 m (Holtmeier 1965, 1974). 

In the Upper Engadine, artifacts were found that give 
evidence of agriculttire during the Bronze Age (about 
3,000 years B.P.), while in the Lower Engadine hunters 
and shepherds were already present in the Neohthic Age 
(4,000 years ago; Conrad 1940). Trading and mining be- 
came increasingly important to the economy dvuing his- 
toric time. Thus, there was increasing human impact on 
the moimtain ecosystems throughout history. Almost no 
untouched nature was left, although the kind and the in- 
tensity of the himaan disturbances were different. 

DEPRESSION OF TIMBERLINE 

Due to the great elevation of the area humans could 
only use the mountains up to a certain altitude; there the 




Figure 1— Location of the Upper Engadine. 



9 



3000 



2500 



2000 



1500 



1000 - 



500 - 



mm 




Figure 2 — Annual precipitation. Note the de- 
crease from Maloja Pass toward St.. Moritz and 
Schuls (Lower Engadine). 





Figure 4 — Zuoz Village (1 ,700 m). The original 
forest was almost totally removed as a conse- 
quence of alpine pasturing (upper slope) and 
agriculture (artificial terraces above the village). 
Only two small stands were left on the steep 
north-exposed slopes of the narrow tributary 
valleys and one above the village for protection 
from avalanches. Photo taken 1 968. 




main influence was concentrated in the forest belt not 
only on favorable southern exposures but also on all 
sections that were fairly accessible — trough shoulders, 
terraces, and similar almost-level areas even in north- 
exposed locations. During former times cereals were 
grown up to elevations of about 2,300 m, favorable topog- 
raphy provided. Many of the high-altitude forests were 
cleared for pastoral use, while those at lower elevations 



gave way to agriculture. The extent of timberline depres- 
sion varied in relation to exposure and topography. On 
some easily accessible southern exposures of the main 
valley, for example, and of the Fex Valley and Fedoz Val- 
ley, the forest was totally removed except for sections 
characterized by too great steepness (figs. 4, 5). Other for- 
ests were destroyed by ore mining, salt works, and char- 
coal production, especially during the Middle Ages. On 




Figure 3 — View from Lej dais Chods (northwest- 
facing slope of the main valley) southwest to- 
ward the gap of Maloja Pass. 




Figure 5 — View to southeast-exposed slope of the main 
valley. In the foreground, Samaden Village (1 ,720 m). 
Forest was removed from all gentle topography (trough 
shoulders, lower slope area) and reduced to a narrow 
belt on the steep trough wall. Photo taken 1969. 



10 




Figure 6 — View from Diavolezza trail to 
Bernina Pass (2,328 m). Bernina Pass was 
covered by larch-cembra pine forest until the 
Middle Ages. Then, the forest became a 
victim of mining. Until present, the area is 
used in summer for grazing cattle. 



the Bernina, Abula, and Jiilier Passes (about 2,300 m) the 
forest disappeared as a consequence of mining. These ar- 
eas v^^ere then used as alpine pastures (fig. 6). 

The upper timberUne receded by 150-300 m compared 
to its position during the postglacial thermal optimvmi. 
That timberline depression was partly due to a general 
cooling. However, at least since the Middle Ages the ef- 
fects of the deterioration of the climate on subalpine tree 
growth have been compounded or even superimposed by 
the human influences mentioned earlier. 

COMPOSITION AND STRUCTURE 

As a consequence of these activities not only did the up- 
per timberline become lower by about 150-300 m, but also 
species composition and structures of the moimtain for- 
ests changed considerably (Auer 1947; Campell 1944; 
Holtmeier 1965, 1967). Larch (Larix decidua) and cembra 
pine {Pinus cembra) are the most important tree species 
in the Upper Engadine forests. These forests belong to 
the so-called silicate type (Larici Cembretum, Ellenberg 
1978; Larici-Pinetum cembrae, Oberdorfer 1970) of larch- 
cembra pine forest, which is common in the crystalline 
Central Alps. The present distribution pattern of both 
species in the Upper Engadine can be described as fol- 
lows: Cembra pine is mainly spread on steep and inacces- 
sible slopes and north exposures, whereas larch domi- 
nates on easily accessible locations exposed to the south. 

The influence of the local topography on the distribu- 
tion pattern is most pronovmced in the trough-shaped 
Bever Valley and Bernina Valley (fig. 7). Pure larch 
stands are confined to the relatively gentle slope areas 
(talus cones, screes, boulder fans, etc.) that extend below 
the steep and rocky trough walls, while cembra pine pre- 
vails on trough walls and also forms the upper timberline. 
This pattern is most conspicuous on southern exposures. 
On north-exposed slopes cembra pine also occurs on their 
lower parts, mixed vnth larch. 



This distribution of larch and cembra pine is mainly 
caused by human influences, especially by pastoral use. 
Since cembra pine impedes the growth of grass and herbs 
on the forest floor, while larch does not, cembra pine was 
cleared on slopes suitable for grazing and restricted to 
steep, inaccessible locations. Moreover, the wood of cem- 
bra pine was used for many purposes such as paneling 
and making furniture, milk vessels, or carvings. Cembra 
pine also is more susceptible to forest fires than larch, 
which is protected by a thick corklike bark. As indicated 
by charcoal -rich soil horizons, forest fires frequently 
occvured in the past. Probably, many of them were 
caused by shepherds when they burned alpine pastures 
to remove "weeds" such as Rhododendron ferrugineum 
and Juniperus nana, and fires ran out of control. Thus, 
for many reasons, larch co;ild spread at the cost of cembra 
pine, as was also the case in many other central-alpine 
valleys. At many places the high-elevation forest is over- 
mature and very sensitive to environmental impacts such 
as windthrow and snow slides. 

CURRENT CHANGES 

Due to modem change of economic structure — tourism 
has become the main base of the economy in the 
Engadine — grazing pressure on the forests declined, and 
cembra pine is invading the former pasture forest. Larch, 
which is a light-demanding species, is going to be out- 
competed by cembra pine, and natural succession ft-om 
larch to cembra pine forests, which had been interrupted 
by human disturbances for some hundred years, is re- 
vived (Holtmeier 1967, 1990, 1993). Larch will gradually 
be replaced by cembra pine, if sites are not specially man- 
aged by exposing mineral soil or selective cutting of 
cembra pine. 

Locally, however, invasion of cembra pine in the imder- 
story is hampered by mass outbreaks of the larch bud 
moth (Zeiraphera diniana). After having defoliated the 
larch crowns, the caterpillars will feed on the cembra 
pines in the understory and, together with secondary 
parasites, cause severe damage to them. Young cembra 
pines become crippled or killed. 

In forest gaps smaller than the height of the surround- 
ing trees, accvunulation of snow and length of v^dnter snow 
cover may critically increase (fig. 8), and consequently 
seedlings and saplings of cembra pine may be seriously 
damaged by snow fungus {Phacidium infestans). 

Cembra pine also is invading abandoned or only rarely 
used alpine pastures. Seeds are dispersed by the nut- 
cracker {Nucifraga caryocatactes), occasionally far beyond 
the present forest limit (Holtmeier 1966, 1974; Mattes 
1978, 1982). Larch seeds, however, are prevented from 
getting into a suitable seedbed by the dense alpine vegeta- 
tion. Thus, it is cembra pine that is the pioneer species 
on former pasture areas (fig. 9), while larch only occurs 
where bare mineral soil is exposed. Survival of the invad- 
ing trees mainly depends on microclimates and resulting 
site conditions. They considerably deteriorated when the 
forest was removed in the past. Favorable effects of the 
general warming on tree growth are superimposed by 
unsuitable local site conditions (fig. 10). Consequently, 
natural reforestation follows the favorable sites while 
other sites will remain treeless probably for a long time. 



11 





Figure 7 — Distribution of tree species in the Bernina Valley (top) and 
Bever Valley (bottom). Both maps have been redrawn from Holtmeier 
(1967). 



12 




Figure 8 — Big snow masses are accumulated 
in the glade. Incoming solar radiation is inter- 
cepted by the trees, while energy loss by out- 
going long-wave radiation is almost unimpeded 
Due to these conditions snowmelt is consider- 
ably delayed, and young cembra pines may be 
infected by snow fungus. 



As to successful germination and survival of seedlings, 
seed dispersal by the nutcracker is much more effective 
than wind-mediated transport because many seeds are 
cached at sites relatively favorable to germination and 
seedling growth (not too long covered with snow). On the 
other hand, the seed caches usually provide sufficient soil 
moisture because they are 2-4 cm under the surface, thus 
being protected from direct insolation. Finally, seeds col- 
lected by the nutcracker are characterized by a relatively 
high germination capacity (Mattes 1978, 1982), and there 
is almost no loss of cached seeds by seed-eating predators. 




Figure 9 — Cembra pine invading an abandoned 
alpine pasture (about 2,250 m) on the northwest- 
exposed slope of the main valley. Invasion fol- 
lows mainly convex topography, while sites 
covered too long with snow and characterized by 
snow creep and avalanches remain treeless. 



However, due to high density of seedling clusters some 
of the trees will become victims of root competition. More- 
over, snow accumulation and duration of the snow cover 
may gradually be increased by the influence of the grow- 
ing trees on windflow near the surface. At such sites, 
yoimg cembra pines covered too long by snow may be 
killed by snow fungi such as Phacidium infestans 
(Holtmeier 1967, 1974, 1990 ). 

Natural invasion by trees (cembra pine, larch, vallows, 
rowan, and also spruce) can be observed in front of and on 
the lateral moraines of the Morteratsch Glacier and Roseg 
Glacier (Holtmeier 1965, 1974, 1990; fig. 11). These gla- 
ciers have been retreating more or less continuously by 
about 30 m per year since the beginning of our century. 
Larch alreadj occurs about 20 years after glacier retreat, 
while cembra pine ynW follow later. Then, however, 
cembra pine invades these areas much more rapidly than 
larch. There is some evidence that areas which became 
ice-free about 50 to 70 years ago were more rapidly in- 
vaded by cembra pine than those from where the glacier 
retreated more than 100 years ago. That might be ex- 
plained by mesoclimatic conditions improving due to gla- 
cier retreat. 



7 




1. Situation during the postglacial thermal optimum 



forest 1 


ecotone 1 alpine 


- 2100 




- 2100 ijpr 


forest climate replaced by the very locally 
changing pattern of microclimates , similar 
to that above the previous tinfterline 


2. Situation at 


present 



Figure 10 — Change of microclimatic conditions in the 
upper Subalpine after the uppermost forest was re- 
moved by humans. 



13 




1.5 km 1 



^Jk cembrapine ^^^i larch ''iV' alpine vegetation 

Figure 11 — Invasion of the forefield and lateral moraine of the retreating Morteratsch Glacier by cembra 
pine and larch (schematically). 



At places, however, distribution patterns of invading 
larches and cembra pines very likely depend on soil mois- 
ture available, as can be observed in the forefield of 
Morteratsch Glacier, for example (fig. 11). Obviously, 
relatively dry sites — such as the well-drained upper part 
of the lateral moraines or roche moutonnee areas — can be 
invaded by cembra pine more easily than by larch, be- 
cause seeds of cembra pine are cached by the nutcracker 
in small cracks or fine-soil pockets between the boulders 
where sufficient moisture is available for germination and 
seedling growth. On the other hand, it depends totally on 
chance whether wind-mediated larch seeds will get into 
such a seedbed. At the valley bottom, conditions appear 
to be more favorable for invasion by larch because of 
great numbers of open patches of bare mineral soil cre- 
ated by extensive accumulation of fine material between 
the boiilders. At such sites soil moisture is usually avail- 
able, and there is a better chance for larch seeds to reach 
a suitable seedbed. 



Spruce (Picea abies) also occurs within cembra pine- 
larch forest. Natural stands, however, are confined to the 
northwest-exposed slope of the main valley between the 
village of Sils and the small Lake of Staz (fig. 12). This 
area is characterized by increased humidity due to a cloud 
belt that usually appears when the Upper Engadine is in- 
fluenced by moisture-carrying air currents from the 
southwest (Holtmeier 1965, 1966, 1967). The upper dis- 
tribution limit of spruce is located at about 2,000 m. Al- 
though spruce is missing between Maloja Pass (1,815 m) 
and Sils, there is some evidence that this species immi- 
grated from the south across Maloja Pass in 6000 to 5500 
B.C. (Campell 1944; Kleiber 1974). The gap can probably 
be attributed to human influences. On the opposite slope 
spruce does not occur, except for planted trees (Schlatter 
1935). That might be explained by lower humidity and 
lower soil moisture on that slope exposed to the southeast. 
Anyway, humans also could have removed shade-giving 
spruce (Holtmeier 1967). After deglaciation, cembra pine 
invaded the Upper Engadine earlier than spruce (Kleiber 




Figure 12 — Distribution of spruce {Picea abies) on the northwest-exposed slope of the main valley (from 
Holtmeier 1967, modified). 



14 




Figure 13 — Arboreous and dwarfed mountain pine in 
bog near Maloja Pass. 



1974). Otherwise it would have been outcompeted by the 
latter species — at least at lower elevations (Heitz 1975; 
Mattes 1978). 

KRUMMHOLZ SPECIES 

Dwarf pine {Pinus mugo) and green alder (Alnus viri- 
dis) form dense and locally extended krummholz stands. 
The krummholz growth of both these species is genetically 
predetermined and not climatically induced. That makes 
them different from climatically trimmed, flagged, and 
matlike "krummholz" that forms the forest-alpine timdra 
ecotone in the Rocky Moimtains, for example (Holtmeier 
1973, 1974, 1981). 

Moreover, there is no contiguous Pinus mugo krumm- 
holz belt above the high-stemmed forest as is peculiar to 
the northern Alps. In the Upper Engadine, both Pinus 
mugo and Alnus viridis occur mainly on slopes endan- 
gered by avalanches. Their elasticity enables them to 
resist avalanches and snow slides better than upright- 
growing tree species, which normally will be eliminated 
at such sites. Some slopes, where the original forest had 
been destroyed by avalanches and fire, were completely 



afforested with moimtain pine. Moimtain pine is also 
common in bogs where it grows on slightly convex topog- 
raphy (mostly rocky outcrops). At such sites the arbore- 
ous growth form also occiu-s, as is the case on Maloja Pass 
or in the Lake Staz area, for example (fig. 13). 

In general, Pinus mugo prevails on limestone and dolo- 
mite, while Alnus viridis is common mainly on silicate 
material. In view of this distribution, both species are 
said to be vicarious plants. In the Upper Engadine, how- 
ever, the availability of soil moistiu^e seems to be the fac- 
tor controlling the distribution pattern. There, Pinus 
mugo is more common on southern exposures, whereas 
Alnus viridis is mainly spread on north-facing slopes pro- 
viding sufficient soil moistiire and fi'esh soils rich in nutri- 
ents. On svmny slopes, however, Alnus viridis is confined 
to moist microsites, as alongside rivlets, for example. On 
northern exposures Pinus mugo successfully competes 
with Alnus viridis at sites relatively dry and poor in nu- 
trients (fig. 14). 

Birch {Betula pubescens) is another species that is usu- 
ally confined to avalanche tracks and bogs. Birch is able 
to recover fi-om breakage by thriving basal shoots. There- 
fore, multistemmed growth is quite common to this spe- 
cies (fig. 15). 

At present, the Upper Engadine has a higher forest 
cover compared to 150 years ago and earlier. Since then 
many afforestations were carried out to close the gaps 
caused by historical hiunan impact. Also "exotic" species 
were used — such as Picea pungens, Picea engelmannii, Pi- 
nus sibirica, and others (Schlatter 1935). It has become 
evident, however, that in this area the native Pinus cem- 
bra and Larix decidua are the most successful species. 

CONCLUDING REMARKS 

The Upper Engadine belongs to the most important 
natural ranges of cembra pine-larch forests in the Alps. 
However, present distribution, composition, and struc- 
tures of these forests can only be explained by permanent 
utilization that began in prehistoric time. That makes 
these forests much different from almost imtouched high- 
altitude forests in remote areas of the Rocky Mountains, 
for example. 

Since tourism has become the main economic base, while 
agriculture and pastoral land use declined, utilization of 
the forest also changed. In our day it is not timber produc- 
tion but establishment and maintenance of a vigorous 
shelter forest (protection from avalanches, soil erosion, 
torrential washes, etc.) that is the main objective of forest 
management. Locally, grazing of the high-altitude forest 
by cattle and game may run counter to the objective and 
should be reduced as much as possible. 

However, the forest also is of great importance for tour- 
ism. Thus, larch, which has a high esthetic value because 
of its bright autvmin colors, is favored by special manage- 
ment practices interrupting natural succession. On the 
other hand, lower elevation forest is still used for produc- 
tion of merchantable timber. Swiss stone pine is used in 
particular for paneling, carvings, and making furniture of 
typical local design. 



15 






PINUS MUCO 




LARIX DECIDUA 


9 <* 



PINUS CEMBRA 



Figure 14 — Distribution of Pinus mugo and AInus viridis on the northeast-exposed slope of the Bernina vaiiey. 
Pinus mugo is confined to relatively dry convex topography, while AInus viridis grows on the moist sites (from 
Holtmeier 1967). 



REFERENCES 

Auer, Christian. 1947. Untersuchungen iiber die natiir- 
liche Verjiingung der Larche im Arven-Larchenwald 
des Oberengadins. Mitteilungen der Schweizerischen 
Anstalt fiir das forstliche Versuchswesen. 25: 3-140. 

Campell, Eduard. 1944. Der Wald des Oberengadins im 
Wandel der Zeiten. Festschrift 124. Jahresversamm- 
lung der Schweizerischen Naturforschenden 
Gesellschaft: 93-102. 

Conrad, H. 1940. Beitrag zur Frage der urgeschichtHchen 
Besiedlung des Engadins. Jahresbericht Historisch- 
Antiquarische Gesellschaft Graubixndens. 70. 
Jahresbericht: 1-40. 




Figure 15 — Birch {Betula pubescens) growing 
in an avalanche track. 



Ellenberg, Heinz. 1978. Vegetation Mitteleuropas mit den 
Alpen in okologischer Sicht. 2d ed. Stuttgart: Verlag 
Eugen Ulmer. 981 p. 

Heitz, Christian. 1975. Vegetationsentwicklung und 
Waldgrenzschwankungen des Spatglazials iind Post- 
glazials im Oberhalbstein (Graubiinden/Schweiz). 
Biindnerwald. 28(2): 45-52. 

Holtmeier, Friedrich-Karl. 1965. Die Waldgrenze im 
Oberengadin in ihrer physiognomischen und okolo- 
gischen DifFerenzierung. University of Bonn, FRG. 
163 p. Dissertation. 

Holtmeier, Friedrich-Karl. 1966a. Die Malojaschlange 
und die Verbreitung der Fichte. Beobachtungen zur 
Klimaokologie des Oberengadins. Wetter und Leben. 
18: 105-108. 

Holtmeier, Friedrich-Karl. 1966b. Die okologische Funk- 
tion des Tannenhahers im Zirben-Larchenwald und an 
der Waldgrenze im Oberengadin. Journal fur Omi- 
thologie. 4: 253-260. 

Holtmeier, Friedrich-Karl. 1967. Die Verbreitung der 
Holzarten im Oberengadin unter dem Einflui3 des 
Menschen und des Lokalklimas. Erdkunde. 21: 249-258. 

Holtmeier, Friedrich-Karl. 1973. Geoecological aspects of 
timberlines in Northern and Central Europe. Arctic and 
Alpine Research. 5(3): 45-54. 

Holtmeier, Friedrich-Karl. 1974. Geookologische Beobach- 
tungen und Studien an der subarktischen und alpinen 
Waldgrenze in vergleichender Sicht (nordliches Fenno- 
skandien/Zentralalpen). Erdwissenschaftliche For- 
schung. 8. 130 p. 

Holtmeier, Friedrich-Karl. 1981. What does the term 
"Krummholz" really mean? Observations with special 
reference to the Alps and the Colorado Front Range. 
Mountain Research and Development. 1(3/4): 253-260. 

Holtmeier, Friedrich-Karl. 1990. Disturbance and man- 
agement in larch-cembra pine forests in Europe. In: 



16 



Schmidt, Wyman C; McDonald, Kathy J., comps. Pro- 
ceedings — symposium on whitebark pine ecoystems: 
ecology and management of a high-momitain resoiirce. 
Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Department 
of Agricultvire, Forest Service, Intermoxantain Research 
Station: 25-36. 

Holtmeier, Friedrich-Karl. 1993. European larch in 
Middle Europe. In: Proceedings — symposium on ecology 
and management of larix forests: a look ahead. Ogden, 
UT: U.S. Department of Agriculture, Forest Service, In- 
termountain Research Station. [In preparation]. 

Kleiber, Helga. 1974. Pollenanalytische Untersuchvmgen 
zum Eisriickgang und zur Vegetationsgeschichte im 
Oberengadin I. Botanisches Jahrbuch fiir Systematik. 
94(1): 1-53. 



Mattes, Hermann. 1978. Der Tannenhaher im Engadin — 

Studien zu seiner okologischen Funktion im Arvenwald. 

Miinstersche Geographische Arbeiten. 2. 87 p. 
Mattes, Hermann. 1982. Die Lebensgemeinschaft von 

Tannenhaher und Arve. Eidgenossische Anstalt fur das 

forstliche Versuchswesen. Berichte. 24. 74 p. 
Oberdorfer, Erich. 1970. Pflanzensoziologische Exkur- 

sionsflora fiir Siiddeutschland. 3d ed. Stuttgart: Verlag 

Eugen Ulmer. 987 p. 
Schlatter, A. J. 1935. Die Aufforstungen und Verbau- 

xmgen des Oberengadins in den Jahren 1875-1934. 

Schweizerische Zeitschrift fur Forstwesen. 86: 309-328. 



17 



Evolution and 
Taxonomy 




International Workshop 
St. Moritz 1 992 

S 



18 



GENETIC DIFFERENTIATION AND 
PHYLOGENY OF STONE PINE 
SPECIES BASED ON ISOZYME LOCI 

Konstantin V. Krutovskii 
Dmitri V. Politov 
Yuri P. Altukhov 



Abstract— F-statistics analysis based on isozyme loci data has 
revealed comparatively low levels of intraspecific genetic differen- 
tiation among natural populations of Eurasian stone pine species: 
Siberian stone pine {Pinus sibirica Du Tour), Korean stone pine 
{Pinus koraiensis Siebold et Zucc), and Japanese (dwarf Siberian 
or mountain) stone pine {Pinus pumila [Pall.] Kegel). Only about 
2 to 4 percent of the total isozyme gene diversity was due to inter- 
population variation (FgT = 0.021-0.040), and the overwhelming 
part of the total variation, over 96 percent, belonged to intrapopu- 
lation variation. These data are discussed in relation to distin- 
guishing features of conifer biology, in general, and to a specific 
mode of stone pine seed distribution by nutcrackers, in particular. 
Cluster analysis of more-studied Siberian stone pine populations 
has shown that genetic distances based on isozyme allele frequen- 
cies reflect geographical distribution of these populations in spite 
of the generally low values of genetic distances (D = 0.001-0.031, 
on the average of 0.006). 

The authors have studied phylogenetic relationships between all 
stone pine species (subsection Cembrae), including one European 
stone pine population— Swiss stone pine {Pinus cembra L.) and one 
population of the only representative of stone pine species in North 
America— whitebark pine {Pinus albicaulis Engelm.), using genetic 
distances based on 16 isozjmie loci, whose genetic control has been 
resolved for all stone pine species. The results have confirmed 
stone pine {Cembrae) species to be a compact group of very closely 
related and genetically similar species, having monophyletic ori- 
gin, supposedly from ancient Siberian stone pine. The smallest 
distance (D = 0.065) has been found between Pinus cembra and 
Pinus sibirica that corresponded, mostly, to their subspecies taxo- 
nomic status. Other interspecific distances were in good agree- 
ment with the species status (D = 0.121-0.268). 

As to the controversy on the taxonomy of whitebark pine, this 
species undoubtedly belongs to subsection Cembrae (D = 0.121- 
0.256). The results of phylogenetic study have been confirmed by 
the recently obtained data of cpDNA restriction fragment analyses. 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our ICnowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Konstantin V. Krutovskii is Senior Research Scientist, Laboratory of 
Population Genetics, N.I. Vavilov Institute of General Genetics, Russian 
Academy of Sciences, GSP-1 Moscow 117809 B-333, Russia (current ad- 
dress until 1995: Department of Forest Genetics and Tree Improvement, 
Georg-August-University, Biisgenweg 2, Gottingen, 37077, Germany); 
Dmitri V. Politov is Research Scientist, same laboratory; Yuri P. Altukhov 
is Professor, Member of Russian Academy of Sciences, Head of the Labora- 
tory of Population Genetics, Director of the Institute of General Genetics. 



Five stone pine species are traditionally united in sub- 
section Cembrae of section Strobus, subgenus Strobus, 
genus Pinus (Critchfield and Little 1966; Shaw 1914). 
European or Swiss stone pine {Pinus cembra) is distrib- 
uted in the Alps and Carpathians and has the smallest 
area among stone pine species. An east Asian stone pine 
species, Korean pine (P. koraiensis), is widely distributed 
in eastern Russia (Russian "Far East"), Manchuria, and 
Japan, but mostly in Korea. Two other stone pine species 
are: Japanese stone pine {P. pumila), widely distributed 
in northeastern Siberia extending to eastern Asia, as far 
as central Japan; and Siberian stone pine (P. sibirica), 
widely distributed, mostly in Siberia extending to northern 
Mongolia and to the eastern border of Europe. Whitebark 
pine (P. albicaulis) is widely distributed in western North 
America, extending from British Coliunbia to the southern 
Sierra Nevada, and represents the only stone pine species 
growing outside Eurasia. 

All these species are of great interest for different kinds 
of investigations, but some main aspects make these species 
very attractive from evolutionary and population genetics 
points of view and explain the necessity of their genetic 
study. 

First, they occupy a vast territory in Eurasia and North 
America and most of them belong to the main forest-forming 
species. Thus, they have great ecological and economic sig- 
nificance, especially in Eurasia. They are subjected to inten- 
sive harvesting and, consequently, need special breeding 
and gene conservation programs based on genetic data. 

Second, stone pine species consist of different kinds of pop- 
ulations. For example, Swiss stone pine includes mostly 
small, isolated populations and occupies a relatively small 
territory. On the other hand, Siberian stone pine includes 
huge, continuous populations occupying a very leirge terri- 
tory in Siberia. Some of the stone pine species are com- 
pletely isolated from others: Swiss stone pine, growing in 
Europe, and whitebark pine, growing in North America, 
stand apart from Asian stone pine species. At the same 
time, Siberian, Korean, and Japanese stone pines have 
sympatric zones. Thus, from evolutionary and population 
genetics points of view and for gene conservation programs 
it would be interesting and important to study genetic dif- 
ferentiation, gene dispersal patterns, and mating system 
and population genetic structure of these species and com- 
pare them with each other. 

Third, in spite of the fact that stone pines have a wide 
geographical distribution and some morphological and 
ecological differences, they are grouped into a separate 
subsection Cembrae, because they have some diagnostic 



19 



morphological traits that distinguish them from other 
Strobus species. Most significant of these are wingless large 
seeds and indehiscent cones that remain closed in matu- 
rity. These traits are supposed to be the result of long co- 
evolution of stone pines and nutcrackers (Nucifraga ssp.), 
who play a crucial role in natural reproduction and dis- 
tribution of stone pine species (Lanner 1980, 1982, 1990; 
Tomback 1983; Tomback and Linhart 1990; Tomback and 
others 1990). Nevertheless, taxonomy of these species is 
still being discussed. Some specialists consider Swiss stone 
pine a subspecies of Siberian stone pine (or vice versa) and 
some think that whitebark pine should be included into 
subsection Strobi because of its morphological and ecologi- 
cal resemblance to some Strobi species (Critchfield 1986; 
Millar and Kinloch 1991). Taking all this into account, the 
use of genetic markers for study of phylogenetic relation- 
ships between stone pine species would be very appropriate 
and fruitful. 

Thus, the main tasks of our research were estimation of 
genetic differentiation and subdivision of stone pine popula- 
tions and study of phylogenetic relationships between stone 
pine species using isozyme loci as genetic markers. 

The advantages and disadvantages of using isozymes in 
plant systematics and phylogeny have been fairly thoroughly 
discussed in scientific literature (see for example, Crawford 
1983; Gottheb 1977; Strauss and others 1992). We would 
only like to note here such taxonomically most important 
advantages as, first, the possibility of using the same trait 
(isozyme electrophoretic pattern encoded by homologous 
loci) to classify taxons of almost any phylogenetic level, both 
related and unrelated, and second, the possibility of quan- 
titative estimation of divergence degree. One of the prob- 
lems of traditional systematics is associated with the objec- 
tive choice of diagnostic traits, but initial selection of such 
traits usually leads to their preliminary "weighting" (evalu- 
ation) and inevitably produces subjectivity. It is, therefore, 
extremely important to use, above all, such traits in taxo- 
nomic studies as can be taken for analysis without "weigh- 
ing," more or less arbitrarily, and about whose variability 
nothing has been known a priori. Molecular genetic mark- 
ers (isozyme loci and DNA markers) are one of the best, 
from these points of view. 

The study presented here is the first one devoted to study 
of phylogeny of the whole subsection Cembrae using isozyme 
markers and is only a part of long-term studies on popula- 
tion genetics of stone pine species carried out in the Labora- 
tory of Population Genetics of N. I. Vavilov Institute of Gen- 
eral Genetics (Russian Academy of Sciences, Moscow, Russia) 
on the basis of isozyme analysis. More details on inheritance 
and linkage of isozjnne loci, mating system, population ge- 
netic structure, and other aspects of stone pine population 
genetics can be found in our previously published papers 
(Krutovskii and Politov 1992; Krutovskii and others 1987, 
1988, 1989, 1990; PoHtov 1989; Politov and Krutovskii 1990; 
Politov and others 1989, 1992) and in the accompanying pa- 
per included in these proceedings by Politov and Krutovskii. 

MATERIAL AND SAMPLING SITES 

Seventeen single-tree and two bulk seed samples from 
natural populations have been used for isozyme analysis in 
our study. Eleven localities of Siberian stone pine, three 



localities each of Korean and Japanese stone pines, and 
one locality each of Swiss stone and whitebark pines have 
been studied on 16-20 isoz3niie loci (table 1). More details 
about seed samples and characteristics of the sampled 
natural populations can be found in our other publications 
(Krutovskii and Politov 1992; Krutovskii and others 1988, 
1989, 1990; Pohtov 1989; Pohtov and Krutovskii 1993; 
Pohtov and others 1992). 

ELECTROPHORETIC ANALYSIS 

Isozymes of seeds have been studied by electrophoresis 
in starch gel. Method of electrophoretic analysis, specimen 
preparation, the buffer systems used, genetic interpretation 
of zymograms, designations of allozymes, alleles, and loci, 
and inheritance of enzyme systems have been described 
elsewhere (Krutovskii and others 1987; Politov 1989). Geno- 
types of trees have been determined by analyzing segrega- 
tion of allozyme alleles among their endosperms, which are 
haploid in conifers and genetically identical to maternal 
tree gametes. Together with adult tree allele frequencies, 
pollen allele frequencies inferred from effective pollen gene 
pool (paternal alleles of the analyzed seed embryos) have 
been used for estimation of genetic distances. These fre- 
quencies of successful pollen give probably more represen- 
tative estimation of the studied populations' gene pools due 
to predominantly outcrossing pollination among stone pine 
species (Pohtov and Krutovskii 1990, 1993). 

In total, 536 trees fi-om 17 localities (11-75 trees each) and 
two bulk seed samples of five stone pine species have been 
analyzed on 16-20 isozyme loci. More than 3,500 (7-20 per 
tree) embryos have been analyzed on a smaller amount of 
isoz5Tne loci because of poor resolution of some enzyme sys- 
tems in embryo tissue. A relatively small number of trees 
have been analyzed in some populations and species, but 
according to some theoretical (Nei 1987a,b) and experimen- 
tal investigations (Shurkhal and others 1992), the number 
of loci used in interspecific phylogenetic analysis is more 
important than the sample size. We have studied phyloge- 
netic relationships between all stone pine species using ge- 
netic distances based on 16 isozyme loci, Adh-1, Dia-2, Fe-2, 
Gdh, (iot-1, Got-2, Got-3, Idh, Mdh-1, Mdh-2, Mdh-3, Mdh-4, 
Pgi-2, Pgm-1, Pgm-2, and Skdh-1, whose allelic variation has 
been unequivocally determined for all stone pine species. 
Such a number of loci is used commonly and gives quite 
reasonable phylogenetic trees (Nei 1987a,b). 

DATA ANALYSIS 

Determination of parameters of intra- and interpopula- 
tional genetic diversity, estimation of genetic differentia- 
tion and genetic distances, and clustering and construction 
of dendrograms for stone pine populations have been car- 
ried out by using IBM PC version 1.7 of the computer pro- 
gram BIOSYS-1 (Swofford and Selander 1981) and IBM 
PC-program RESTSITE, version 1.1 (Nei and Miller 1990). 

Using our data, we have estimated different kinds of ge- 
netic distances and tried different methods of clustering. 
Methods for calculating genetic distances and methods of 
clustering differ in their assumptions concerning homoge- 
neity of evolutionary rates or proportionality to the time of 
divergence. Consequently, topology of derived phylogenetic 



20 



Table 1 — Name, abbreviated designation, sample size, and geographic location of sampling sites 



Locality 



Designation 



Number 
of trees 



Region 



Siberian Stone Pine 



Filin Klyuch 



Mutnaya Rechka M 

Maiyi Kebezh MK 

Sobach'ya Rechka SR 

Listvyanka L 

Yailyu YA 

Smokotino S 

Zorkartsevo-1 Z-1 

Zorkartsevo-3 Z-3 

Turukhansk TU 



Nyalino NYA 



Ust'-Chornaya UCH 



Grustnyi GR 
Uttyveem UT 
Kamenskoye K 



Malokhekhtsirskoye MKH 
Khor KHO 
Sikhote-Alin' SA 



Livingstone Falls LF 



53 



75 
15 
12 
41 
43 

43 
34 
34 

MOO 



Western Sayan Mountains (Ermakovskii 
District, Krasnoyarsk Territory, 
eastern Siberia) 



Altai Mountains (Gorno-Altaiskaya 

Autonomous Region, eastern Siberia) 
Tomskaya Region (western Siberia) 



'88 



North of Krasnoyarsk Territory, 
(Turukhanskii District, North of 
eastern Siberia) 
Khanty-Mansi Autonomous District 
(Tyumenskaya Region, North of 
western Siberia) 

Swiss Stone Pine 

15 Eastern Carpathians (western Ukraine) 

Japanese Stone Pine 

54 North of Kamchatka Peninsula 

55 

55 



Korean Stone Pine 



30 
16 
11 

Whitebark Pine 

52 



Khabarovsk Territory 

(Russian "Far East") 

Primorskii Territory (Russian "Far East") 



Rocky Mountains Forest Reserve, 
Bow-Crow Forest (southwestern 
Alberta, Canada) 



'Number of analyzed seeds from mixed sample obtained from large number (>100) of trees. 



trees based on different distances and methods of cluster- 
ing can vary significantly for the same data. 

Among methods of clustering that assimie homogeneity 
of evolutionary rates, an unweighted pair group method 
using arithmetic averages, UPGMA, is the most widely 
used one (Sneath and Sokal 1973). Among methods that 
do not assume homogeneity of evolutionary rates, the dis- 
tance Wagner procedure (Farris 1972) and neighbor-joining 
method, NJM (Saitou and Nei 1987) are the most widely 
used ones. 

The most common genetic distances used in isozjnne 
studies of trees are Nei's (Nei 1972, 1978), Rogers' (1972), 
and CavalU-Sforza and Edwards' (1967) distances. How- 
ever, Nei's distance, D, along with advantages has some 
disadvantages, the main one being that it is not metric 
because it does not obey the triangle inequality and, conse- 
quently, cannot be used in the distance Wagner procedure 
or NJM. Thus, we have used D (Nei 1972) only in UPGMA 
dendrograms. 



Taking into accoimt: (1) that homogeneity of evolution- 
ary (or allozyme allele substitution) rates can usually be 
safely assumed in comparing conspecific populations of 
widespread, continuous tree species with high gene flow, 
like stone pine populations; (2) shortage of space for pre- 
sentation of all obtained dendrograms; and (3) possibility 
for wider comparison vnth published data, we will present 
only UPGMA dendrograms based on Nei's distance, D, 
for analysis of intraspecies stone pine population relation- 
ships. However, to avoid problems connected with possible 
nonhomogeneity of evolutionary rates of different species, 
we will present NJM dendrograms of stone pine species 
based on Cavalli-Sforza and Edwards' (1967) chord dis- 
tances, Dch, in addition to UPGMA dendrograms based on 
D and D.^. 

Nei's genetic distances matrix has also been used for 
principal coordinate analysis of Siberian stone pine popula- 
tions using the computer program NTSYS-pc (Rohlf 1988), 



21 



INTRASPECIFIC DIFFERENTIATION 

Levels of Genetic Differentiation Between 
Populations — Distribution of intraspecific genetic variation 
between populations of stone pines has been analyzed for 
three species studied in several localities using F-statistics 
(Nei 1977). Levels of genetic differentiation among popula- 
tions were comparatively low — only about 2-4 percent of 
the total intraspecific isozyme gene variation was due to 
interpopulation variation (FgT = 0.021-0.040), and the over- 
whelming part of the total variation, over 96 percent, be- 
longed to intrapopulation variation (table 2). Loci Fe-2, 
Mdh-3, and Skdh-1 of Siberian stone pine, Adh-1, Adh-2, 
Gdh, Got-2, Mdh-2, Mdh-3, Mdh-4, and Skdh-2 of Japanese 
stone pine, and Mdh-4 of Korean stone pine make a most 
significant contribution in intraspecific differentiation of 
these species (table 2). The geographic variation of these 
loci may be adaptive and should be studied in detail in fu- 
ture investigations. 

Nei's genetic distances between populations were also 
small, D = 0.001-0.031 and 0.012, on the average (table 3). 
These values correspond perfectly to the calculations car- 
ried out using references on 22 pine species (Altukhov and 
others 1989). The genetic distances between closely located 
(<20 km) populations of seven pine species equaled 0.009, 
on the average, and between widely dispersed populations 
of 15 pine species reviewed — D = 0.036, on average. 

Low levels of interpopulation differentiation of allozyme 
loci are usual for conifers. Such typical factors for most coni- 
fer populations as outcrossing, wind pollination, seed dis- 
persal by wind (most conifers) or by birds (stone pines, some 
white pines, and pinyons), wide continuous ranges, high 
population density, and effective size are considered to re- 
duce the influence of genetic drift and, therefore, decrease 



heterogeneity of allele frequencies and interpopulation 
genetic differentiation (Hamrick 1983; Hamrick and Godt 
1989; Hamrick and Loveless 1986; Hamrick and others 
1981; Loveless and Hamrick 1984). 

We have also simmiarized data on interpopulation genetic 
differentiation of 28 conifer species of eight genera estimated 
using FgT or analogous GgT parameters of differentiation 
(Politov and others 1992). In spite of a, generally, compar- 
atively low differentiation among populations, we have 
found a quite pronounced tendency to a significantly smaller 
differentiation (Fst = 0.02-0.08) among widespread species 
with continuous population ranges than among species with 
interrupted ranges and small, and often isolated, popula- 
tions (Fg-r = 0.11-0.13). Moreover, according to our review, 
samples from the same population (samples fi'om localities 
within several kilometers radius) revealed, commonly, an 
even much lower level of differentiation (Fg^ = 0.006-0.02). 

The three stone pine species studied belong to the first 
group of species with mainly continuous population range. 
Besides, four of the 11 Siberian stone pine samples, F, M, 
MK, and SR, have been collected within 20 km radius from 
obviously the same population of Malyi Kebezh River Basin 
(eastern Siberia, table 1); and the other two samples have 
been taken fi^om two stands near Zorkal'tsevo village (west- 
em Siberia), Z-1 and Z-3, which are separated from each 
other by less than 3 km. The other localities are rather re- 
mote from one another, but all of them represent only the 
central, continuous part of the species area. There are no 
significant barriers to gene flow in this area. 

The low differentiation observed in our study may also 
be due to the limited number of samples taken, mostly, 
from more or less optimal zones of the stone pines area, 
impeding the diversifying selection action and promoting 
the balancing one (Altukhov 1990). 



Table 2 — The estimates of F-statistics for populations of three stone pine species and for all five species 



Locus 


Siberian stone pine 




Korean stone pine 




Japanese stone pine 




All 




Ft 






Ft 


Fst 


F,s 


Ft 


Fst 


Fst 


Adh-1 


-0.030 


-0.006 


0.023 


0.078 


0.120 


0.045 


0.306 


0.321 ' 


0.022 


'0.563 


Adh-2 


{') 


{') 


{') 


(') 




{') 


-.068 


-.046 


\020 




Dia-1 


-.032 


-.007 


.024 


(') 




(') 


{') 


n 


{') 




Dia-2 


-.060 


-.053 


.006 


{') 


{') 


(') 


-.038 


-.013 


\025 


'.028 


Fe-2 


-.061 


-.012 


\046 


-.067 


-.021 


.043 


-.105 


-.097 


.007 


'.190 


Gdh 


-.012 


-.001 


.010 


-.098 


-.098 


.000 


.084 


.140 


^061 


'.637 


Got-1 


-.013 


-.001 


.012 


{') 


{') 


{') 


{') 


{') 


{') 


.002 


Got-2 


-.012 


-.001 


.010 


{') 


e) 




-.050 


-.022 


\027 


'.016 


Got-3 


(') 


{') 


{') 


.238 


.241 


.004 


.014 


.015 


.001 


'.596 


Lap-2 


.015 


.025 


.009 


-.071 


-.023 


.045 


-.048 


-.047 


.001 




Lap-3 


.066 


.079 


.013 


(') 


{') 


{') 


-.064 


-.050 


.013 




Mdh-2 


.214 


.231 


.023 


-.251 


-.225 


.020 


-.004 


.015 


^019 


'.240 


Mdh-3 


-.049 


.000 


\047 


-.067 


-.021 


.043 


-.051 


.084 


M28 


'.118 


Mdh-4 


-.068 


-.055 


.012 


.017 


.104 


'.089 


-.021 


-.013 


\ooa 


'.362 


Pgi-2 


.114 


.123 


.010 


e) 


{') 


{') 


-.111 


-.107 


.003 


'.407 


Pgm-1 


-.074 


.051 


.022 


-.193 


-.160 


.028 


.177 


.189 


.016 


'.187 


Pgm-2 


{') 


(') 


{') 


{') 


?) 




-.014 


-.013 


.002 


'.046 


Skdh-1 


-.146 


-.118 


\025 


-.316 


-.194 


.092 


-.053 


-.051 


.002 


'.187 


Skdh-2 


{') 


{') 


{') 


{') 


{') 




.099 


.163 


'.071 




Mean 


-.043 


-.018 


.025 


-.077 


-.033 


.040 


-.002 


.020 


.021 


.396 



'This locus has statistically significant heterogeneity of allele frequencies among populations. 

n"he genetic base of isozyme variation of this locus is not determined or allelic variation cannot be compared with other species, 
^his locus is monomorphic in this species. 



22 



Table 3 — Genetic distances, D, (Nei 1972) for five stone pine species 







Sibarian 






Ivor oan 


Species 




stone pine 


stone pine 


stone pine 


Stone pine 


Siberian stone pine 


11 


0.006 
(0.002-0.016) 








Swii^^ stonp oinp 


"1 


0 065 
(0.055-0.077) 








Jaoanese stone oine 


3 


0.202 
(0.176-0.253) 


0.233 
(0.215-0.265) 


0 010 
(0.003-0.014) 




Korean stone pine 


3 


0.235 


0.268 


0.105 


0.009 






(0.207-0.261) 


(0.264-0.275) 


(0.096-0.119) 


(0.004-0.014) 


Whitebark pine 


1 


0.121 


0.130 


0.185 


0.256 






(0.111-0.129) 




(0.166-0.220) 


(0.253-0.261) 



'N = number of localities studied. 



And finally, the low differentiation between stone pine 
populations can be considered the result of a specific mode 
of seed dispersal by nutcrackers able to carry their seeds 
over very long distances (up to 20 km, according to some 
observations). The efficiency of this mode and its influence 
on the population structure were repeatedly emphasized 
by researchers studying relationships between nutcrackers 
and whitebark pine, Swiss stone pine, limber pine {Pinus 
flexilis James), and pinyons (subsection Cembroides) 
(Carsey and Tomback 1992; Furnier and others 1987; 
Lanner 1980, 1990; Schuster and Mitton 1991; Tomback 
and others 1990, 1992). 

Cluster and Principal Coordinates Analysis — 

Despite the low values of genetic distances, UPGMA cluster- 
ing of stone pine samples based on these distances reflects, 
in general, their geographical origin. More detailed analy- 
ses have been performed on Siberian stone pine, which has 
been studied more than other stone pines. We can tenta- 
tively divide all 11 Siberian stone pine localities into three 
regions, according to their ecology and geographic location. 
The first of them can be called "Southern Siberia Moimtains" 
and includes one population from the Altai Mountains (Ya) 
and five localities in the western Sayan (FK, M, MK, SR, 
and L); the second can be called "Western Siberia Plain" 
and includes populations from Tomsk region (S, Z-1, and 
Z-3); the third, "Northern Siberia," includes populations 
from Khanty-Mansi Autonomous District (NYa) and from 
the north of Krasnoyarsk Territory (TU). The dendrogram 
obtained corresponds exactly to this classification (fig. 1). 
These three regions form three different genetic groups 
(clusters) of populations. 

Korean pine samples also form clusters in accordance 
with their population localization (two geographically more 
closely spaced stands, MKh and Kho, have turned out to 
be genetically more similar). The same is true of Japanese 
stone pine populations (fig. 1). Principal coordinates analy- 
ses based on the matrix of Nei's genetic distances between 
Siberian stone pine populations also have been performed. 
This multidimensional analysis has produced almost the 
same result as the cluster analysis. Genetic relationships 
between Siberian stone pine populations, graphically pre- 
sented in two-dimensional space of the first two principal 
coordinates, have perfectly reflected the real geographical 



distribution of these populations (fig. 2). Principal coordi- 
nates and cluster analyses complement each other and have 
produced almost the same result in our case, but the first, 
probably, gives more information about population genetic 
relationships due to better data presentation. Multivariate 
analysis techniques are supposed to provide, basically, bet- 
ter possibility for revealing genetic differentiation in coni- 
fers, as compared to univariate methods (Krutovskii and 
Politov 1992; Yeh and others 1985). 

Statistically significant positive correlation between geo- 
graphical and genetic distances has been found for Siberian 
stone pine populations (r = 0.844, P < 0.001). Such correla- 
tion was, however, absent for samples from closely spaced 
localities of the Malyi Kebezh River Basin. The limited 
number of samples studied has not allowed such analysis 
for other stone pine species, but analogous results have 



0.010 



0.005 
I 




D 




F 




L 
M 
MK 


Southern 

Siberia 

Mountains 


YA 




SR _ 




Z-1 

Z-3 


Western 

Siberia 

Plain 


NYA _ 


Northern 
Siberia 


MKH~ 
KHO_ 


1 Khabarovsk" 
Territory 


SA 


Primorski! - 
Territory 


GR 
UT 

K - 


Kamchatka 
Peninsula 



Siberian 

stone pine 



Korean 
stone pine 

Japanese 
stone pine 



Figure 1 — UPGMA phenograms of stone pine 
populations based on Nei's genetic distances, 
D, (Nei 1972). (See full names and origins of 
populations in table 1 .) 



23 



NYA 



Southern 
Siberia 
Mountains 




Western 
Siberia Plain 



-0.15 



Figure 2 — Projections of Siberian stone pine 
populations in two-dimensional space of the first 
two principal coordinates, X and Y, according to 
Nei's genetic distances, D, (Nei 1972). (See full 
names and origins of populations in table 1 .) 



also been obtained for some other pine species (Furnier 
and Adams 1986; Steinhoff and others 1983; Wheeler and 
Guries 1982), including, probably, that isolation by distance 
(Nei 1975), introgressive hybridization (Millar 1983), or 
gradiently changing selection (Bergmann 1978) could play 
a significant role in genetic differentiation of some pine 
species. 

We have also analyzed correlation between geographical 
and Nei's distances calculated for Siberian stone pine popu- 
lations using data on individual loci (allele frequencies of 
eight loci in the effective pollen pool have been used). For 
loci Adh-1, Dia-2, Fe-2, Pgm-1, and Skdh-1 correlation was 
significant. The other loci (Lap-3, Mdh-2, and Pgi-2) dis- 
played lack of correlation. This may suggest a selective role 
of certain loci. Patterns of spatial differentiations seem to 
be determined by complex interaction of gene flow and selec- 
tion. Distribution of genetic variability may be also affected 
by "evolutionary footprints," such as, for example, Siberian 
population descent from different refugiums. 

INTERSPECIFIC DIFFERENTIATION 

Genetic Distances Between Stone Pine Species — 

We have estimated both genetic distances, D^j, and D, be- 
tween all stone pine species based on 16 isozyme loci data, 
but only D values are presented in table 3, because of space 
limits. The smallest genetic distances have been found 
between Siberian and Swiss stone pines (D = 0.052-0.077). 
They are just a little larger than those between subspecies 
of lodgepole pine (Pinus contorta Dougl. ex Loud.), D = 0.008- 
0.019, approximately 0.012 on the average (Wheeler and 
others 1983), similar to the distances between subspecies 
of brutian pine {Pinus brutia Ten.), D = 0.03-0.09 (Conkle 
and others 1988), and a little smaller than distances be- 
tween two very closely related species, jack pine (Pinus 
banksiana Lamb.) and lodgepole pine, which have a wide 
zone of introgressive hybridization, D = 0.097, according to 



the study by Dancik and Yeh (1983), or D = 0.108, according 
to Wheeler and Guries (1987). Nei's distances estimated 
between other pairs of stone pine species (D = 0.121-0.268, 
table 3) are close to the values reported for pine species 
of subsections Contortae: 0.078-0.158 (Wheeler and others 
1983) and Oocarpae: 0.11-0.27 (Millar and others 1988). 
Approximately the same value, D = 0.27-0.43, has been 
reported for Pinus brutia-Pinus halepensis comparison 
(Conkle and others 1988). We have estimated the average 
genetic distance between species fi-om the same subsections 
using a few references (Conkle and others 1988; Millar and 
others 1988; and some others). It equaled 0.185, on the av- 
erage, for 14 pine species and surprisingly corresponded to 
the average distance estimated for all stone pine species in 
our study — D = 0.180, and for four Eurasian stone pine spe- 
cies studied by Shurkhal and others (1991a,b)— D = 0.138 
(for comparison, D = 0.185, on the average, for four Eurasian 
stone pines used in our study). The small discrepancy be- 
tween the latter two estimates is not surprising, taking into 
account the sampling strategy used by Shurkhal and others 
(1991a,b; 1992). They have analyzed only a few trees of 
each species (mostly, from Botanical Garden collections), 
which should inevitably bias genetic distance estimates to 
smaller values, due to a higher probability of random fixa- 
tion of most common, frequent alleles in small samples of 
trees from different species. 

Levels of interspecific genetic differentiation obtained 
between stone pines strongly confirm their taxonomical 
status as closely related species of the same subsection. 
Nei's genetic distances obtained for species belonging to 
different subsections or sections are usually higher than 
0.3 and vary between 0.3 and 0.8 (Shurkhal and others 
1991a,b; 1992). According to phylogenetic analysis carried 
out by Shiu-khal and others (1991a,b; 1992) using isoz)Tne 
loci, fotir stone pine species of subsection Cembrae (white- 
bark pine has not been analyzed) and three white pine spe- 
cies of subsection Strobi, Balkan or Macedonian pine (Pinus 
peuce Griseb.), blue or Himalayan pine (Pinus griffithii 
McCleland), and eastern white or We3maouth pine (Pinus 
strobus L.), form separate clusters. The Nei's genetic dis- 
tance between these two very closely related subsections 
equals 0.278, on the average (calculated by us, using the 
data in table 5 from Shurkhal and others 1991b). This is 
almost two times larger than that between Eurasian stone 
pines (D = 0.138). 

The lack of species-specific diagnostic loci among stone 
pines also confirms the correctness of uniting them in the 
same separate subsection Cembrae. Only Siberian and 
Korean stone pines, on locus Gdh, and Swiss and Korean 
stone pines, on locus Mdh-4, do not have common alleles. 
Besides, Korean stone pines have a unique electrophoretic 
pattern of LAP enzyme indicating the existence (or expres- 
sion) of an additional Lap-4 locus. 

Loci Adh-1, Gdh, Got-3, Mdh-4, and Pgi-2 make a most 
significant contribution to interspecific differentiation of 
stone pines, but totally only 39 percent of the whole sub- 
section Cembrae genetic variation belongs to interspecific 
differences (table 2). 

Thus, our data give a strong genetic evidence of common, 
apparently monophyletic stone pines origin and compara- 
tively recent time of their divergence. Subsection Cembrae 
seems to be a valid taxon. 



24 



Phylogenetic Relationships Inferred From Genetic 
Distances — Many publications have been devoted to stud- 
ies of taxonomic and phylogenetic relationships between 
stone pines using morphological, physiological, ecological, 
and other traits. Our study was the first one to use genetic 
markers, isozyme loci, for analysis of phylogenetic relation- 
ships between all five stone pine species. These species are 
traditionally united in subsection Cembrae, according to 
some characteristic features (Critchfield and Little 1966; 
Farjon 1984; Little and Critchfield 1969; Shaw 1914). The 
overwhelming majority of researchers consider the main spe- 
cific features of this subsection, indehiscent cones and wing- 
less seeds, to be inherited fi*om a common ancestor. These 
traits seem to be a relatively recent acquisition. Their ap- 
pearance is considered to be a result of coevolution of stone 
pines and nutcrackers of genus Nucifraga foraging their 
seeds and playing the main role in their maintenance and 
dispersal (Lanner 1980, 1982, 1990; Tomback 1983; 
Tomback and Linhart 1990; Tomback and others 1990). 

Thus, the subsection Cembrae is regarded as a natural 
monophyletic taxon. Nevertheless, there also exists a dif- 
ferent point of view (Critchfield 1986), according to which 
traits shared by all stone pines are insufficiently confirmed 
as homologous. Data on artificial hybridization appear to 
be rather contradictory (Critchfield 1986). The subsection 
is believed to have polyphyletic origin and, therefore, is 
not a valid taxon (see for discussion, Lanner 1990). Millar 
and Kinloch (1991) also have pointed out a nimiber of im- 
solved problems in taxonomy and phylogeny of subsection 
Cembrae. 

Molecular and biochemical genetic markers have proven 
to be a useful tool for study of biosystematic and phyloge- 
netic relationships in conifers (Strauss and others 1992). 
We believe that application of biochemical genetic markers 
like isozymes can help shed light on the evolution of stone 
pine species. The dendrograms produced by NJM and 
UPGMA clustering using D^h and D matrix (table 3) based 
on 16 isozyme loci have apparently revealed real phyloge- 
netic relationships between five stone pine species (fig. 3). 

According to this dendrogram, Siberian and Swiss stone 
pines appear to be a most closely related pair of species 
(D = 0.065). This was not imexpected, because of the well- 
known surprisingly high morphological similarity (Bobrov 
1978) and successful artificial hybridization (Critchfield 
1986) between them. Practically all traditional botanical 
classifications note the absence of diagnostic morphological 
traits, which means that interspecific differences are more 
quantitative than qualitative. According to paleobotanical 
(mainly palynological) data, Bobrov (1978) considered these 
two species as two now separately located marginal parts 
of a former single ancestral species spread on a vast range 
area, from the western Alps to the northeastern Yakutia 
(eastern Siberia), whose pollen has been foimd in Paleocene, 
Pleistocene, and later deposits. 

As Bobrov (1978) indicates, "Siberian stone pine must be 
regarded as the initial, more ancient species, formerly more 
widely distributed, whereas the West European stone pine, 
Pinus cembra, is evolutionarily still very yoimg and has be- 
come geographically isolated in the recent past, being, there- 
fore, hardly disting\iishable from its Siberian ancestors." 

Breakup of this ancient range, its division into two parts, 
and decrease of its western part occurred, most likely, as 



UPGMA 



Isozyme loci 



L 



JL 



-L 



225 200 175 150 125 100 75 50 25 



UPGMA 



cpDNA RFLP 



Whitebark pine 
Swiss stone pine 
Siberian stone pine 
Korean stone pine 
Japanese stone pine 

D (X10-3) 



Whitebark pine 
Swiss stone pine 
Siberian stone pine 
Korean stone pine 
Japanese stone pine 



J. 



± 



± 



J. 



90 80 70 60 50 40 30 20 10 



UPGMA 



isozyme loci 



C3 



d (X10-5) 



Whitebark pine 
Swiss stone pine 
Siberian stone pine 
Korean stone pine 
Japanese stone pine 



_L 



_L 



_L 



I 



405 360 315 270 225 180 135 90 45 



Dch(x10-3) 



NJM isozyme loci 

Siberian stone pine 



C Siberian 



± 



± 



Swiss stone pine 

Whitebari< pine 

— — Korean stone pine 
-■i-v— i— i^^— Japanese stone pine 

-J \ \ lD,,(x10-3) 



45 90 135 180 225 270 315 360 405 



NJM 



Swiss stone pine 
Siberian stone pine 

Korean stone pine 




cpDNA RFLP 

Whitebark pine 



I L 



J L 



15 



30 



45 



65 



Japanese stone pine 
-Jd (X10-5) 



Figure 3 — UPGMA and NJM dendrograms of five 
stone pine species (subsection Cembrae) based on 
Cavalli-Sforza and Edwards' (1967) and Nei's (Nei 
1972) genetic distances, and D, calculated from 
isozyme data, and on d (Nei 1 987b; Nei and Miller 
1990), calculated from chloroplast (cp) DNA restric- 
tion fragment length polymorphism (RFLP) data 
(Krutovskii and Wagner 1993). 



25 



late as in the Holocene and was associated with its substi- 
tution in the central part of the area by other species due 
to the global warming. The time of divergence (10,000 to 
25,000 years), estimated from the genetic distance between 
Siberian and Swiss stone pines corresponds, approximately, 
to these events (for more details see Krutovskii and others 
1990). Thus, the low level of divergence between Siberian 
and Swiss stone pines can be explained by the short time 
of isolation. 

Nevertheless, a considerable amount of genetic difference 
has accumulated during this time — genetic distances be- 
tween Siberian and Swiss stone pines are approximately 
10 times larger than those between populations of the same 
species (table 3) and are comparable with distances reported 
by Conkle and others (1988) for subspecies and very closely 
related pine species. Our data have also been recently con- 
firmed by Goncharenko and others (1992) and Shurkhal and 
others (1991a,b; 1992). Their estimates of Nei's genetic dis- 
tance between Siberian and Swiss stone pines equaled 
0.030 and 0.042, correspondingly. Thus, Swiss stone pine 
seems to be a highly diverged geographical race or subspe- 
cies of Siberian stone pine, or a species in statu nascendi. 

Unexpectedly relatively small differentiation found be- 
tween Japanese and Korean stone pines (D = 0.105) makes 
us search for a new approach to understanding their origin. 
These species, in fact, have little in common in morphology 
of both vegetative and generative organs, but according to 
the study of parenchjnnal tissue of needle cell walls, stone 
pine species and some closely related species can be divided 
into two t3T)es: Siberian and Swiss stone pine form the first 
type, while Japanese and Korean stone pines belong to the 
second type, but to different subtypes (Litvintseva 1974). 
Perhaps, the evolutionary pathways of these species have 
truly separated from a common ancestor relatively recently. 
Another explanation of their genetic similarity is gene ex- 
change through introgressive hybridization, which could 
have occurred (or seldom occurs), taking into account the 
present sympatry of these species and, as a rule, poorly 
developed mechanisms of reproduction isolation between 
closely related pine species. More detailed discussion can 
be foimd in Krutovskii and others (1990), but recently ob- 
tained data also confirm the division of Eurasian stone pines 
by isozyme loci analysis into two groups: Siberian and Swiss 
stone pines (D = 0.030-0.042), on one hand, and Korean and 
Japanese (D = 0.089-0.143), on the other hand (Goncharenko 
and others 1992; Shurkhal and others 1991a,b; 1992). 

A most interesting problem, from our point of view, is ex- 
planation of the whitebark pine's closer affinity to Siberian 
and Swiss stone pines (D = 0.121-0.130) than to Japanese 
stone pine (D = 0.185), although quite opposite relation- 
ships could be suspected considering these species' present 
area. Among stone pines Siberian and Swiss stone pines 
are, probably, genetically most similar to the original an- 
cient stone pine, due to large refugiums in the Alps and 
Siberian (Altai and Sayan) Mountains, which could help 
them save a great deal of ancient genetic variation; and 
whitebark pine also could keep more genetic similarity with 
the ancient species than Korean and Japanese stone pines. 
Another explanation could be that stabilizing selection sup- 
ports genetic similarity between whitebark pine and Siberian 
and Swiss stone pines, because whitebark pine ecologically 



has, probably, more in common with them than with Korean 
and Japanese pines. 

However, the dendrogram obtained by NJM, and based 
on Cavalli-Sforza and Edwards' (1967) distance, D^^, has 
revealed a kind of "intermediate" position of whitebark pine 
between two stone pine groups, Swiss and Siberian stone 
pines, on one hand, and Korean and Japanese stone pines, 
on the other (fig. 3). In these cases NJM gives more an in- 
tuitively realistic phylogenetic tree. It is important to note 
that both UPGMA dendrograms based on D^,, and Nei's dis- 
tance, D, have practically the same topology, but they differ 
from the NJM dendrogram. Thus, one can conclude that 
assumption of equal evolutionary rates of stone pine spe- 
cies can be erroneous. 

Millar and Kinloch (1991) discuss several theoretically 
possible variants of whitebark pine origin: (1) from ancestral 
Eurasian stone pine with penetration into North America 
during the Mesozoic (before Laurasia segregation). Ter- 
tiary, or Quartemary (through the Bering land bridge) time; 
(2) fix)m one of the American pines of subsection Strobi. Now 
we can finally solve this problem. Nei's genetic distances 
(D) between whitebark pine and Eurasian stone pines equal 
approximately 0.121-0.256 (table 3). The time of divergence 
it) between whitebark pine and Eurasian stone pines, theo- 
retically estimated from these values, is 0.6-1.3 million years, 
using calibration coefficient (k) 5x10^ in the equation t = kD 
(Nei 1975; 1987a,b). This corresponds quite well to the ap- 
proximate geological time of the Bering Strait opening — 
1.8-3.5 million years ago — estimated by different methods 
(see references in Grant 1987). Consequently, this event, 
the Bering Strait opening, can be, obviously, considered 
as the beginning of genetic differentiation between North 
American and Eurasian pines due to geographical isolation. 

Thus, the hypothesis of a relatively recent penetration of 
ancient stone pines into North America through the Bering 
lemd bridge in the Pliocene seems most preferable. If we 
assumed the hypothesis that a more ancient penetration 
of stone pines through a "Euramerican" connection had 
occurred during Mesozoic time (Millar and Kinloch 1991), 
much more differentiation between whitebark pine and 
Eurasian stone pines should have been found in our study. 
The assumption of whitebark pine origin from Strobi pines 
must be rejected, according to our data, but phylogenetic 
relationships of stone pines with other relative species from 
subsections Strobi and Cembroides certainly need much 
more analysis. The first steps in this direction have been 
made recently using cpDNA markers. 

Chloroplast DNA Data and Pine Phylogeny— Only 
a few studies on pine phylogeny using DNA markers have 
been made. This allows us to draw a short comparison of 
the Millar and Kinloch (1991) results with our data, pre- 
sented here, paying special attention to phylogenetic rela- 
tionships of stone pines with some other closely related 
species. 

Chloroplast DNA (cpDNA) restriction fragment length 
polymorphism (RFLP) of the five stone pine species and four 
white pine species of a closely related subsection Strobi — 
limber pine, sugar pine (P. lambertiana Dougl.), western 
white pine (P. monticola Dougl.), and eastern white pine — 
have been recently studied by Krutovskii and Wagner (1993). 
Data have been obtained by molecular hybridization of 27 



26 



probes, representing the entire chloroplast genome of 
P. contorta, with restriction fragments from 20 endonu- 
cleases. The total nimiber of restriction fragments per 
sample varied from 378 to 422. 

Chloroplast DNA variation among trees within species 
was not pronounced (D = 0.0-0.0013) and was comparable 
with variation between populations within species, but 
cpDNA diversity, in general, was quite high — 50 to 100 
percent individual cpDNA belonged to different haplotypes 
(for more details see Krutovskii and Wagner 1993). The 
dendrogram constructed on Nei's genetic distances (D), 
estimated from these data (Nei 1987b; Nei and Miller 
1990), has shown almost the same phylogenetic relation- 
ships between stone pine species as established on isozyme 
loci data (fig. 3). The stone and white pine species have 
been imdoubtedly divided into two separate groups (fig. 4). 
It is interesting to note that NJM dendrograms give better 
correspondence between cpDNA and isozyme data. Com- 
paring NJM and UPGMA dendrograms we can suppose 
that assumption of homogeneity of evolutionary rates is 
less proper for cpDNA than for nuclear (isozyme) genes. 

The cpDNA data also indicate that genetic differentia- 
tion is more developed among the four Strobi species than 



among the Cembrae species. The same concliision has been 
made by Shurkhal and others 1991a,b; 1992) on the isozyme 
data. 

Chloroplast DNA's of whitebark pine and Eurasian stone 
pine species were remarkably similar. This similarity, to- 
gether with clear cpDNA divergence between the two sub- 
sections, confirms systematic classification of whitebark 
pine in subsection Cembrae, and not in subsection Strobi. 
Phylogenetic relationships of whitebark pine with Siberian 
and Swiss stone pines revealed by cpDNA RFLP analysis 
are surprisingly consistent with the conclusions from the 
analysis of 16 isozyme loci (fig. 3). Thus, ovir observation 
is not an artifact, because two independent sets of genetic 
markers (isozymes and cpDNA RFLPs) have produced the 
same conclusions. Moreover, Szmidt (1991) notes that the 
present geographic distribution of some forest tree species 
often shows little correlation with the degree of cpDNA dif- 
ferentiation among them, because the present geographic 
distribution of some species is of relatively recent origin. 

The data obtained by Szmidt and others (1988), who used 
four restriction enzymes for cpDNA RFLP analysis of 12 
pine species, including two stone pine species (Swiss stone 
and whitebark pines) and two white pine species (western 



UPGMA 



cpDNA RFLP 



Whitebark pine 
Swiss stone pine 
Siberian stone pine 
Korean stone pine 
Japanese stone pine 
Western white pine 
Limber pine 
Sugar pine 
Eastern white pine 

d (xlO-3) 



Stone 
Pines 



White 
Pines 



30 



20 



10 



NJM 



J Sit 

it 



Siberian stone pine 
Swiss stone pine 
Whitebark pine 
Korean stone pine 
apanese stone pine 

Western white pine 
Limber pine 
Sugar pine 




Stone 
Pines 



Eastern white pine 
I id(x10-3) 



White 
Pines 



10 



20 



30 



Figure 4 — UPGMA and NJM dendrograms of five stone pine spe- 
cies (subsection Cembrae) and four white pine species closely re- 
lated to them: Pious flexilis, P. lambertiana, P. monticola, and P. 
strobus (subsection Strobi), based on Nei's genetic distances, d, 
(Nei 1987b; Nei and Miller 1990), and calculated from chloroplast 
(cp) DNA restriction fragment length polymorphism (RFLP) data 
(Krutovskii and Wagner 1993). 



27 



and eastern white pines), also confirm our results. They 
evaluated phylogenetic relationships between these species 
using UPGMA clustering based on dissimilarity index, anal- 
ogous to Nei's D. According to the dendrogram given by 
Szmidt £ind others (1988), Swiss stone and whitebgirk pines 
form a very tight separate cluster. Western white pine is 
distantly linked to this cluster, which exactly corresponds 
to our data (fig. 4). Eastern white pine forms a separate 
branch but is closely linked to the group of Swiss stone, 
whitebark, and western white pines. 

Analogous data have been obtained by Strauss and 
Doerksen (1990) who studied one sample fi"om each of 18 
to 19 pine species representing 14 to 15 subsections using 
eight restriction enzjnnes and 17 cloned DNA fragments 
(mostly fi"om Pseudotsuga menziesii). At least 65 percent 
of the chloroplast genome was covered by the combination 
of the probes used. The samples have also included two 
stone pine species (Korean stone and whitebark pines) and 
two white pine species (blue and sugar pines). Distances 
between the samples were estimated via shared characters. 
The stone £uid white pine species were extremely similar 
to one another, with distances ranging fi^om 0.0 to 0.018. 
They form one cluster. The authors have mentioned (data 
have not been presented) that other, more limited, restric- 
tion fi'agment studies of section Strobus, including three 
stone (Swiss, Siberian, and Japanese) and three white pines 
(Armand's [P. armandii Franch.], limber, and western white), 
have given the same indication of high uniformity. Unfor- 
tunately, Strauss and Doerksen (1990) have not been able 
to distinguish stone and white pines, probably because a 
limited number of restriction enzymes or DNA probes have 
been used in their research. 

Certainly, stone pines and other closely related species 
need more genetic analysis using both isozyme and DNA 
markers. Many fascinating aspects of their evolution are 
sure to remain hidden, and many discoveries are still w£dt- 
ing for research. 

CONCLUSIONS 

The main conclusions based on our data are: 

1. Both biochemical (isozymes) and molecular (cpDNA) 
genetic markers are very useful and informative for study 
of population genetic structure and phylogeny of stone pine 
and other closely related species, and they complement each 
other. 

2. In spite of the high level of isozyme loci and cpDNA 
diversity among stone pine species, genetic differentiation 
between the populations is not pronounced, but specific. 
As a minimum, the genetic differentiation reflects their 
geographical and ecological peculiarities. Thus, it is neces- 
sary to study populations fi'om the whole of the stone pine 
species area for careful description of gene distribution, 
taking into account even very small genetic differences 
between populations for developing gene conservation and 
reforestation programs. 

3. According to both isozyme and cpDNA restriction fi'ag- 
ment analyses, P. albicaulis undoubtedly belongs to Cembrae 
species and not to Strobi species. 

4. Stone pine species consist of a compact group of very 
closely related and genetically similar species, obviously 



having monophyletic origin, supposedly, fi*om ancient 
Siberian stone pine. Thus, subsection Cembrae should 
be considered as a valid taxon. 

ACKNOWLEDGMENTS 

We are very gratefiil to V. N. Vorobjev and N. A. Vorobjeva 
(Institute of Ecology of Natural Complexes, Tomsk, Russia), 
L. I. Milyutin and his laboratory staff (Institute of Forest 
and Timber, Krasnoyarsk, Russia), G. R. Furnier (Univer- 
sity of Minnesota, Minneapolis, MN, U.S.A.), R, T. Gut (For- 
est Technology Institute, Lvov, Ukraine), and P. Razumov 
(Far East Forest Institute, Khabarovsk, Russia) for their 
help in collecting and kindly providing Cembrae species seed 
samples for isozjnne analysis. We are th£inkful to Yu. S. 
Belokon' (Institute of General Genetics, Moscow, Russia) 
for very helpfiil assistance in electrophoretic ansdysis, R. M. 
Lanner (Utah State University, Logan, UT, U.S.A.), D. F. 
Tomback (University of Colorado at Denver, Denver, CO, 
U.S.A.), and C. Millar (Institute of Forest Grenetics, Berkeley, 
CA, U.S.A.) for very useful comments on pine phylogeny. 
We also appreciate H.-R. Gregorius' invaluable comments 
that helped to substantially improve the manuscript. 

REFERENCES 

Altukhov, Y. P. 1990. Population genetics: diversity and sta- 
bility. New York: Harwood Academic Publishers. 352 p. 

Altukhov, Y. P.; Krutovskii, K V.; Dukharev, V. A.; 
Larionova, A. Y.; Politov, D. V.; Ryabokon', S. M. 1989. 
Biochemicgd population genetics of woody forest species. 
In: Iroshnikov, A. I., ed. International symposium on for- 
est genetics, breeding and physiology of woody plants: 
proceedings; 1989 September 24-30; Voronezh. Moscow: 
The State Forest Committee of the USSR: 21-29. 

Bergmginn, F. 1978. The allelic distribution of an acid phos- 
phatase locus in Norway spruce {Picea abies) £dong simi- 
lar climatic gradients. Theoretical and Applied Genetics. 
24: 57-64. 

Bobrov, E. G. 1978. Forest-forming conifers of the USSR. 
Moscow: Nauka. 190 p. [In Russian]. 

Carsey, K. S.; Tomback, D. F. 1992. Growth form distribu- 
tion and genetic relationships in tree clusters of Pinus 
flexilis, a bird-dispersed pine. Oecologia. [In press.] 

Cavalli-Sforza, L. L.; Edwards, A. W. F. 1967. Phylogenetic 
analysis: models and estimation procedures. American 
Journal of Human Genetics. 19(2): 233-257. 

Conkle, M. T.; Schiller, G.; Grunwald, C. 1988. Electro- 
phoretic analysis of diversity and phylogeny of Pinus 
brutia £ind closely related taxa. Systematic Botany. 
13(3): 411-424. 

Crawford, D. J. 1983. Phylogenetic and systematic infer- 
ences fi'om electrophoretic studies. In: Tanksley, S. D.; 
Orton, T. J., eds. Isoz)nnes in plant genetics and breed- 
ing. Part A. Amsterdam: Elsevier Scientific Publishers: 
257-287. 

Critchfield, W. B. 1986. Hybridization and classification 
of the white pines (Pinus section Strobus). Taxon. 35(4): 
647-656. 

Critchfield, W. B.; Little, E. L. 1966. Geographic distribu- 
tion of the pines of the world. Washington, DC: U.S. De- 
partment of Agriculture. 99 p. 



28 



Dancik, B. P.; Yeh, F. C. 1983. Allozyme variability and 
evolution of lodgepole pine {Pinus contorta var. latifolia) 
and jack pine (Pinus banksiana) in Alberta. Canadian 
Journal of Genetics and Cytology. 25: 57-64. 

Farjon, A. 1984. Pines: drawings and descriptions of the 
genus Pinus. Leiden: E. J. Brill. 204 p. 

Farris, J. S. 1972. Estimating phylogenetic trees from dis- 
tance matrices. American Naturalist. 106(951): 645-668. 

Fumier, G. R.; Adams, W. T. 1986. Geographic patterns of 
allozyme variation in Jeffrey pine. American Journal of 
Botany. 73(7): 1009-1015. 

Fumier, G. R.; Knowles, P.; Clyde, M. A.; Dancik, B. P. 1987. 
Effect of avian seed dispersal on the genetic structure of 
whitebark pine populations. Evolution. 41(3): 607-612. 

Goncharenko, G. G.; Padutov, V. E.; Silin, A. E. 1992. Popu- 
lation structure, gene diversity, and differentiation in natu- 
ral populations of cedar pines (Pinus subsection Cembrae, 
Pinaceae) in the USSR. Plant Systematics and Evolu- 
tion. 182(3-4): 121-134. 

(rottlieb, L. D. 1977. Electrophoretic evidence and plant 
systematics. Annals of Missouri Botanical Garden. 64: 
161-180. 

Grant, W. S. 1987. Genetic divergence between congeneric 
Atlantic and Pacific Ocean fishes. In: Ryman, N.; Utter, 
F., eds. Population genetics and fishery management. 
Seattle, WA: University of Washington Press: 225-246. 

Hamrick, J. L. 1983. The distribution of genetic variation 
within and among natural plant populations, hi: Chambers, 
S.; Schonewald-Cox, C, eds. Genetics and wild population 
management. Reading, PA: Addison- Wesley: 335-349; 
501-509. 

Hamrick, J. L.; Godt, M. J. 1989. Allozyme diversity in plant 
species. In: Brown, A. H. D.; Clegg, M. T.; Kahler, A. L.; 
Weir, B. S., eds. Plant population genetics breeding and 
genetic resources. Sanderland, MA: Sinauer Associates: 
43-63. 

Hamrick, J. L.; Loveless, M. D. 1986. The influence of seed 
dispersal mechanisms on the genetic structure of plant 
populations. In: Estrada, J.; Fleming, T. H., eds. Frugi- 
vores and seed dispersal. Dordrecht, The Netherlands: 
Dr. W. Junk Publishers: 211-223. 

Hamrick, J. L.; Mitton, J. B.; Linhart, G. B. 1981. Levels of 
genetic variation in trees: influence of life history charac- 
teristics. In: Conkle, M. T., ed. Symposium on isozymes 
in North American forest trees and forest insects: proceed- 
ings. Berkeley, CA: U.S. Department of Agricultiu-e, For- 
est Service: 35-41. 

Krutovskii, K. V.; PoHtov, D. V.; Altukhov, Yu. P. 1987. 
Genetic variability in Siberian stone pine, Pinus sibirica 
Du Tour. I. Genetic control of isozyme systems. Genetika. 
23(12): 2216-2228. [In Russian]. [English translation pub- 
lished by Plenvun Publishing Corp. in Soviet Genetics; 
1988 June: 1568-1576]. 

Krutovskii, K. V.; Pohtov, D. V.; Altukhov, Yu. P. 1988. 
Genetic variability in Siberian stone pine, Pinus sibirica 
Du Tour. II. Level of allozyme variability in western 
Sayan. Genetika. 24(1): 118-124. [In Russian]. [English 
translation published by Plenum Publishing Corp. in 
Soviet Genetics; 1988 July: 81-86]. 

Krutovskii, K V.; Pohtov, D. V.; Altukhov, Yu. P.; Milyutin, 
L. I.; Kuznetsova, G. V.; Iroshnikov, A. I.; Vorobjev, V. N.; 
Vorobjeva, N. A. 1989. Genetic variability in Siberian 
stone pine, Pinus sibirica Du Tour. IV. Genetic diversity 



and differentiation between poptilations. Genetika. 25(11): 
2009-2032. [In Russian]. [EngHsh translation published 
by Plenum Publishing Corp. in Soviet Genetics; 1990 
May: 1343-1362]. 

Krutovskii, K. V.; PoHtov, D. V.; Altukhov, Yu. P. 1990. In- 
terspecific genetic differentiation of Eurasian stone pines 
for isoenzyme loci. Genetika. 26(4): 694-707. [In Russian]. 
[English translation published by Plenum Publishing 
Corp. in Soviet Genetics; 1990 October: 440-450]. 

Krutovskii, K. V.; Politov, D. V. 1992. Study of intra- and 
interspecific genetic differentiation between E\u*asian 
cedar pines using isozyme loci and multidimensional 
analysis. In: Symposium on molecular mechanisms of 
genetic processes: proceedings. Moscow: Nauka Publish- 
ers: 87-96. [In Russian]. 

Krutovskii, K. V.; Wagner, D. B. 1993. [Unpubhshed.] Chlo- 
roplast DNA variation and phylogeny in Pinus subsec- 
tions Cembrae and Strobi (subgenus Strobus). Lexington, 
KY: University of Kentucky, Department of Forestry. 
[Working draft]. 

Lanner, R. M. 1980. Avian seed dispersal as a factor in the 
ecology and evolution of limber and whitebark pines. In: 
6th North American Forest Biology Workshop: proceed- 
ings; 1980 August 11-13; Edmonton, AB. Edmonton, AB: 
University of Alberta: 15-47. 

Lanner, R. M. 1982. Adaptations of whitebark pine for seed 
dispersal by Clark's nutcracker. Canadian Journal of 
Forest Research. 12: 391-402. 

Lanner, R. M. 1990. Biology, taxonomy, evolution, and geog- 
raphy of stone pines of the world. In: Schmidt, Wyman C; 
McDonald, Kathy J., comps. Proceedings — sjonposium 
on whitebark pine ecosystems: ecology and management 
of a high-mountain resource. Gen. Tech. Rep. INT-270. 
Ogden, UT: U.S. Department of Agriculture, Forest Ser- 
vice, Intermountain Research Station: 14-24. 

Litvintseva, M. V. 1974. Characteristics of needle paren- 
chymal cell structure in species of the group Cembrae 
of the genus Pinus. Botanicheski Zhurnal. [Leningrad]. 
59(10): 1501-1505. [In Russian]. 

Little, E. L.; Critchfield, W. B. 1969. Subdivisions of the 
genus Pinas (pines). Res. Rep. 1144. Washington, DC: 
U.S. Department of Agriculture, Forest Service: 1-51. 

Loveless, M. D.; Hamrick, J. L. 1984. Ecological determi- 
nants of genetic structure in plant populations. Annual 
Review of Ecology and Systematics. 15: 65-95. 

Millar, C. I. 1983. A steep cline in Pinus muricata. Evolu- 
tion. 32(2): 311-319. 

Millar, C. I.; Kinloch, B. B. 1991. Taxonomy, phylogeny, and 
coevolution of pines and their stem rusts. In: Hiratsuka, Y.; 
Samoil, J. K.; Blenis, P. V.; Crane, P. E.; Laishley, B. L., 
eds. Rusts of pine. Inf. Rep. NOR-X-317. Edmonton, AB, 
Canada: Northern Forestry Centre. 

Nei, M. 1972. Genetic distance between populations. 
American Naturahst. 106: 283-292. 

Nei, M. 1975. Molecular population genetics and evolution. 
New York: American Elsevier. 288 p. 

Nei, M. 1977. F-statistics and analysis of gene diversity in 
subdivided populations. Annals of Human Genetics. 41: 
225-233. 

Nei, M. 1978. Estimation of average heterozygosity and 
genetic distance from a small niunber of individuals. 
Genetics. 89: 583-590. 



29 



Nei, M. 1987a. Genetic distance and molecular phylogeny. In: 
Ryman, N.; Utter, F., eds. Population genetics and fishery 
management. Seattle, WA: University of Washington 
Press: 193-223. 

Nei, M. 1987b. Molecular evolutionary genetics. New York: 
Columbia University Press. 512 p. 

Nei, M.; Miller, J. C. 1990. A simple method for estimating 
average nvunber of nucleotide substitutions within and 
between populations from restriction data. Genetics. 125: 
873-879. 

Politov, D. V. 1989. Allozyme polymorphism, genetic differ- 
entiation, and mating system in Siberian stone pine, Pinus 
sibirica Du Tour. Moscow: Institute of General Genetics, 
Russian Academy of Sciences. 190 p. Dissertation Thesis. 
[In Russian]. 

Politov, D. v.; Krutovskii, K. V.; Altukhov, Yu. P. 1989. 
Genetic variability in Siberian stone pine, Pinus sibirica 
Du Tour. III. Linkage relationships among isozyme loci. 
Genetika. 25(9): 1606-1618. [In Russian]. [Enghsh trans- 
lation published by Plenum Publishing Corp. in Soviet 
Genetics; 1990 March: 1053-1063]. 

Politov, D. v.; Krutovskii, K. V. 1990. Genetic variability 
in Siberian stone pine, Pinus sibirica Du Tour. V. Analy- 
sis of mating system. Genetika. 26(11): 2002-2011. [In 
Russian]. [English translation published by Plenum Pub- 
Hshing Corp. in Soviet Genetics; 1991 May: 1309-1316]. 

Politov, D. v.; Krutovskii, K V.; Altukhov, Yu. P. 1992. Iso- 
zyme loci characteristics of gene banks of populations 
of cedar pines. Genetika. 28(1): 93-114. [In Russian]. 
[English translation published by Plenum Publishing 
Corp. in Soviet Genetics; 1992 July: 76-95]. 

Rogers, J. S. 1972. Measures of genetic similarity and ge- 
netic distance. In: Studies in Genetics VII. Publ. 7213. 
Austin, TX: University of Texas: 145-153. 

Rohlf, F. J. 1988. NTSYS-pc: numerical taxonomy and mul- 
tivariate analysis system. New York: Exeter Publishing 
Ltd. 

Saitou, N.; Nei, M. 1987. The neighbor-joining method: a 
new method for reconstructing phylogenetic trees. Molecu- 
lar Biology and Evolution. 4: 406-425. 

Schuster, W. S. F.; Mitton, J. B. 1991. Relatedness within 
clusters of bird-dispersed pine and the potential for kin 
interactions. Heredity. 67: 41-48. 

Shaw, G. R. 1914. The genus Pinus. Publ. Arnold Arboretimi 
No. 5. Cambridge, IVLA: Riverside Press. 96 p. 

Shurkhal, A.; Podogas, A.; Zhivotovsky, L. 1991a. Genetic 
differentiation of 18 pine species on allozyme loci, genus 
Pinus, subgenus Strobus and Pinus. Doklady Akademii 
Nauk SSSR. 316(2): 484-488. [In Russian]. [English trans- 
lation published by Plenum Publishing Corp. in Doklady 
Biological Sciences; 1991]. 

Shurkhal, A.; Podogas, A.; Zhivotovsky, L. 1991b. Phyloge- 
netic analysis of genus Pinus by allozyme loci: genetical 
differentiation of subgenus. Genetika. 27(5): 1193-1205. 
[In Russian]. [English translation published by Plenum 
Publishing Corp. in Soviet Genetics; 1991 November]. 

Shurkhal, A.; Podogas, A.; Zhivotovsky, L. 1992. Allozyme 
differentiation in the genus Pinus. Silvae Genetica. 41(2): 
105-109. 

Sneath, P. H. A.; Sokal, R. R. 1973. Numerical taxonomy. 
San Francisco: W. H. Freeman: 230-234. 



Steinhoff, R. J.; Joyce, D. G.; Fins, L. 1983. Isozyme varia- 
tion in Pinus monticola. Canadian Journal of Forest Re- 
search. 13(6): 1122-1131. 

Strauss, S. H.; Doerksen, A. H. 1990. Restriction fragment 
analysis of pine phylogeny. Evolution. 44(4): 1081-1096. 

Strauss, S. H.; Bosquet, J.; Hipkins, V. D.; Hong, Y.-P. 
1992. Biochemical and molecular genetic markers in bio- 
systematic studies of forest trees. New Forests. 6: 125-158. 

Swofford, D. L.; Selander, R. B. 1981. BIOSYS-1: a FOR- 
TRAN program for the comprehensive analysis of elec- 
trophoretic data in population genetics and systematics. 
Journal of Heredity. 72: 281-283. 

Szmidt, A. E. 1991. Phylogenetic and applied studies on 
chloroplast genome in forest conifers. In: Hattemer, H. H.; 
Fineschi, S.; Cannata, F.; Malvolti, E. V., eds. Biochemi- 
cal markers in the population genetics of forest trees. 
The Hague, The Netherlands: SPB Academic Publishing: 
185-196. 

Szmidt, A. E.; Sigurgeirsson, A.; Wang, X. R.; Hdllgren, 
J. E.; Lindgren, D. 1988. Genetic relationships among 
Pinus species based on chloroplast DNA pol3miorphism. 
In: H^lgren, J.-E., ed. Frans Kempe Sympositun in Umeli 
on molecular genetics of forest trees: proceedings; 1988; 
Umed, Sweden. Ume&, Sweden: Swedish University of 
Agricultural Sciences: 33-47. 

Tomback, D. F. 1983. Nutcrackers and pines: coevolution or 
coadaptation? In: Nitecki, N. H., ed. Coevolution. Chicago: 
University of Chicago Press: 179-223. 

Tomback, D. F.; Hoffmann, L. A.; Sund, S. K. 1990. Coevo- 
lution of whitebark pine and nutcrackers: implications for 
forest regeneration. In: Schmidt, Wyman C; McDonald, 
Kathy J., comps. Proceedings — symposiimi on whitebark 
pine ecosystems: ecology and management of a high- 
mountain resource. Gen. Tech. Rep. INT-270. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, Inter- 
movmtain Research Station: 118-129. 

Tomback, D. F.; Holtmeier, F. K.; Mattes, H.; Carsey, K. S.; 
Powell, M. L. 1992. Tree clusters and growth form distri- 
bution in Pinus cembra, a bird-dispersed pine. Denver, 
CO: University of Colorado at Denver, Department of 
Biology. [Working draft]. 

Tomback, D. F.; Linhart, Y. B. 1990. The evolution of bird- 
dispersed pines. Evolutionary Ecology. 4: 185-219. 

Wheeler, N. C; Guries, R. P. 1982. Population structure, 
genetic diversity and morphological variation in Pinus 
contorta Dougl. Canadian Journal of Forest Research. 
12(3): 595-606. 

Wheeler, N. C; Guries, R. P. 1987. A quantitative measure 
of introgression between lodgepole and jack pines. Cana- 
dian Journal of Botany. 65: 1876-1885. 

Wheeler, N. C; Guries, R. P.; O'Malley, D. M. 1983. Bio- 
systematics of the genus Pinus, subsection Contortae. 
Biochemical Systematics and Ecology. 11: 333-340. 

Yeh, F. C. H.; Cheliak, W. M.; Dancik, B. P.; Illingworth, K; 
Trust, D. C; Pryhitka, B. A. 1985. Population differentia- 
tion in lodgepole pine, Pinus contorta ssp. latifolia: a dis- 
criminant analysis of allozyme variation. Canadian Jour- 
nal of Genetics and Cytology. 27(2): 210-218. 



30 



COEVOLUTIONAL ASPECTS OF STONE 
PINES AND NUTCRACKERS 

Hermann Mattes 



Abstract — ^Nutcrackers (Nucifraga caryocatactes and N. 
Columbiana) have developed a strong mutualistic relationship to 
stone pines. Seeds of stone pines are the most important food for 
at least 9 months of the year and for raising the young. In addi- 
tion to special adaptations on gathering, transporting, caching, 
and finding again the hoarded seeds, the whole annual cycle of 
the nutcracker's life (time of breeding and moulting), its mating 
system, and its habitat use are adjusted to the use of pine seeds. 



Within the family of jays, magpies, and crows (Cor- 
vidae) many characteristics of preadaptive value for seed 
dispersal occur (Tomback and Linhart 1990; Turcek and 
Kelso 1968). Three corvid species have evolved a strong 
mutualistic system with pines: Clark's nutcracker (Nuci- 
fraga Columbiana) and the pinyon jay (Gymnorhinus 
cyanocephalus) in North America and the Exiropean nut- 
cracker {Nucifraga caryocatactes) in Eurasia. Compre- 
hensive studies on food regime and seed dispersal were 
carried out by many authors (Bibikov 1948; Crocq 1990; 
Hayashida 1982, these proceedings; Marzlviif and Balda 
1992; Mattes 1978; Reijmers 1959; Saito 1983; Swanberg 
1951; Tomback 1977; Vander Wall and Balda 1977, 1981). 
Coevolution in corvids and pines does not only refer to 
feeding and caching behavior, but the whole life cycle is 
adjusted to the use of the special food resource. This pa- 
per describes the process for the European nutcracker. 

THE EUROPEAN NUTCRACKER 

The annual life cycle (fig. 1) of the nutcracker is best 
looked at fi'om August onward. Unripe seeds of the Swiss 
stone pine (cembra pine, Pinus cemhra) are eaten by the 
nutcracker throughout August or even in July; thus, the 
bird will not miss the point when the seeds are fully de- 
veloped. At this time the resin flow in the cone stops, and 
cone scales get loose. Now the seeds can be easily taken 
out of the cone, and the nutcracker begins gathering and 
caching seeds for winter supply. Pine seed gathering will 
last until all seeds have been harvested and almost no 
cones are left on the trees. Only twice within 19 years of 
investigation in the Engadine a cone crop lasted until the 
beginning of the following spring. 

In the Alps, nutcrackers carry seeds of cembra pine over 
distances up to 15 km (including differences in altitude of 
at least 700 m) back to their territories. They carry 45 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Hermann Mattes is Professor of Landscape Ecology and Geography, 
Department of Geography, University of Miinster, Germany. 



pine seeds per flight on an average; however, a maximum 
load of 134 cembra pine seeds has been found in the sub- 
lingual pouch of a nutcracker. Main morphological and 
ethological adaptations for seed use are: 

• a long and pointed bill (fig. 2) that enables the bird 

to take seeds out of the cones as one would with twee- 
zers (most other corvids have a hooked bill); 

• a rist in the lower mandible allows the bird to fix and 
crack pine seeds, supported by a short-cut tongue; 

• the subhngual pouch allows carrying a large quantity 
of seeds (fig. 3) without disturbing other functions of 
the bill; 

• ability to test seed quality by means of seed color and 
resonance (by so-called bill-clicking). 

SEED CACHES 

The cached seeds provide the main food source for the 
nutcracker from November to April, and to a lesser degree 
until July. A single nutcracker stores more than 25,000 
seeds. As yet it has not been clearly shown how the nut- 
cracker can retrieve its seed caches. However, as can be 
concluded fi'om observations in the field and in the labora- 
tory, nutcrackers apparently remember each single cache 
(Tomback 1982, and others). Further, there are no spe- 
cific features of caches, although nutcrackers obviously 
prefer to deposit seeds near border lines and any changes 
of surface structure. Experiments with the Clark's nut- 
cracker by Kamil and Balda (1988) and Vander Wall 
(1982) showed that nutcrackers orient (navigate) opti- 
cally, as expected, and additionally with a magnetic sense. 
Understanding cache findings, however, requires more de- 
tailed information, as nutcrackers detect caches even un- 
der a thick cover of snow. Balda and Kamil (1989) sug- 
gest an overall fuzzy optical navigation for finding caches. 
This wovild allow the birds to locate a cache even if there 
were distinctive changes, for example, in height of snow 
cover, broken branches, or fallen trees. It is still an open 
question how the bird can find caches under a snow cover 
of 20 cm and higher, getting very directly to the hoarded 
seeds. Snow tunnels of 130 cm (Burckhardt 1958) and 
about 300 cm (R. Stem, personal communication) were 
reported, and — what is most surprising — the nutcrackers 
did find the cached seeds. Crocq (1990) found the nut- 
crackers leaving their subalpine habitat when snow cover 
reached 170 cm on imdistixrbed surfaces. 

It shoiild be emphasized that nutcrackers dig success- 
fully for their caches throughout winter imtil early spring. 
That guarantees sufficient food for the brood during that 
difficult time of the year. Nutcrackers must remember 
not only the sites of the caches but additionally those that 
already have been used. This doubles the efforts of 



31 



> 

twmm 




11 



Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. 

Month 

Figure 1 — Annual activities of the nutcracker in the Alps for years with a medium cone crop. 1 -3, feeding: 
1 , seed harvest and transport; 2, use of cached seeds; 3, consumption of animal food, for example, spi- 
ders, insects; 4, winter period; 5-7 breeding period: 5, courtship and nest building; 6, time with clutch and 
nestlings; 7, family groups with fledged young; 8-9, moulting: 8, moult of adults; 9, moult of fledglings; 
10-11, movements: 10, dismigration; 11, migration. 



memory. There is some evidence that nutcrackers use 
caches from the penultimate seed crop, which may be im- 
portant in case a cone crop fails. In years of small seed 
crops, nutcrackers often recache the seeds. Temporary 
caches are made in cases of strong competition for seeds 
(when great nimabers of nutcrackers and squirrels concen- 
trate on a few cone-bearing trees). In temporary caches, 
nutcrackers mostly hide cones. 

BREEDING AND HABITAT 

Nutcrackers start breeding in March (fig. 4). At this 
time mean daily temperature is under 0 °C. Early breed- 
ing is necessary because the next generation has to be 
ready for harvesting seeds in August to ensure the winter 
food supply. Breeding takes about 18 days, the nestlings 
fledge when 24 to 28 days old, and there is an extraordi- 
narily long time of parental care for the fledglings — until 
they are about 100 days old (Oberauer 1991; Rudat 1978). 
Extensive care by the parents is needed because of scarcity 




Figure 2 — Inside the nutcracker's bill: short 
split tongue, underneath opening of sublin- 
gual pouch with peanut in the entrance part, 
rist in the lower mandible. 



32 




Figure 3 — Nutcracker with filled sublingual 
pouch. Morteratsch, Engadin; September 
1975. 



of food up to mid-June, and to provide training for caching 
and digging for seeds. 

Courtship behavior starts in the end of February or 
early March. Synchronizing with the partner is of ex- 
treme importance for successful breeding. By February 
the nutcrackers' so-called ceremonial gatherings begin 
and then can be frequently observed (Swanberg 1956). 
Up to 12 or more nutcrackers from neighboring and some- 
times far territories meet in conspicuous noisy groups. An 
individually marked female nutcracker that had spent the 
winter outside its territory came back late — March 17. I 
discovered her fighting for the territory while a ceremo- 
nial gathering was going on. She succeeded in reoccupy- 
ing her former place, and raised two nestlings after hav- 
ing a three-egg clutch with one unfertile and half-sized 
egg. Neighbors know each other individually. So, except 
in artifically created feeding places, fights are very rare. 

Nutcracker territories are organized in a different way 
than those of other passerines. Only the area very close 
to the nest is defended against intruders and thus can be 
called "territory" in the usual sense. In a broader sense, 
the territory comprises the area of food caches and is 
about 5 to 12 ha large according to habitat. These "terri- 
tories" can overlap and are not defended because cached 
seeds are only accessible to the very same nutcracker. In 
addition, all places regularly visited for seed gathering 
and transport should be regarded as home range. Habi- 
tats of the nutcracker are characterized by evergreen co- 
nifers. In such forests sheltered nesting places are avail- 
able, and because snow cover is low under dense tree 
canopies cached seeds can be easily recovered. 



Nutcrackers are adapted in many ways for breeding 
under winter conditions. The nest is heavily built with a 
layer of dry decayed wood for insulation against low tem- 
perature. Beginning with the first egg, the parents must 
warm the clutch without long interruptions. Breeding 
starts after the last egg has been laid, and egg tempera- 
ture is raised from about 25 °C to breeding conditions of 
about 38 °C. Different from most corvids, both adults 
breed. The male develops a large breeding patch, too. 
The female's breeding time is somewhat higher during the 
day, and it also stays on the nest at night. The nestlings 
are fed with pine seeds from their first day of life. Addi- 
tional food consists of spiders, and, if available, insects. 
Nutrient analysis in seeds of whitebark pine {Pinus 
albicaulis) by Lanner and Gilbert (these proceedings) 
showed a high nutritive quality in caloric value, mineral 
nutrients, fatty acids, and amino acids. Should this apply 
also to cembra pine seeds, it is easy to understand why 
first-day nestlings can live mainly on a seed diet. 

For successful overwintering and reproduction, suffi- 
cient numbers of seed caches must be available to both 
parents. First-year birds dismigrate in August and estab- 
lish their own territory when seed caching starts. First- 
year birds that are not fully developed until August will 
have a minor chance to survive winter. A few first-year 
birds leave their native area in September and migrate in 
southern directions (Mattes and Jenni 1984). Adult nut- 
crackers are highly sedentary. Nonbreeding individuals 
might have action ranges within or between the territo- 
ries of breeding birds. This popiilation reserve can com- 
pensate for winter losses of adult birds. 

In the "Stazer Wald" I observed a third bird helping a 
pair build their nest; unfortunately, no further observa- 
tion was possible. Similar to several jay species, helping 
in breeding may be an evolutional feature in the nut- 
cracker (Woolfenden and Fitzpatrick 1984). Nutcrackers 
live monogamously for life, as can be concluded from indi- 
vidually marked birds (Mattes 1978; Swanberg 1956). In 




Figure 4 — A nutcracker breeding, the nest 
completely covered by snow. March 1 9, 
1975, Engadin. 



33 



the Engadine, the mortality rate in adult birds was 84 per- 
cent (Mattes, in preparation) on an average. The oldest 
birds found were at least 15. Keeping territories and part- 
ners for life supports synchronization of the parents, and 
individual knowledge of neighbors minimizes social com- 
petition. Social hierarchy is evident and most conspicu- 
ous at artifical feeding places. Females take over the so- 
cial rank of their mates (Grabher-Meyer 1991). 

In passerines, moult normally takes place after the 
breeding season. Prolonged care of the fledglings and 
beginning of seed hoarding in August would not leave 
time for movdting in late summer. Actually, moult starts 
in March with change of remiges and tail feathers, and 
ends in July when body feathers are changed. Moult com- 
pletely parallels breeding season; the double physiological 
effort causes a very long moulting period. Body feathers 
are moulted relatively quickly during the warmest part of 
the year. 

REACTION TO CHANGES 

Predation on pine seeds has influenced the whole life 
cycle of the nutcracker. Many features common to the 
corvids are particularly developed in the nutcracker. Still 
there is a broad and flexible reaction to environmental 
changes. This is most apparent during the invasions of 
Siberian nutcrackers (N. c. macrorhynchos) into western 
Europe. Diet and feeding as well as habitat use vary 
widely. Nevertheless, most of the invading birds very 
likely die. Some, however, return to their breeding ground 
either in the current autumn or in the next spring. Nimaer- 
ous birds ringed in the Aland archipelago (Finland) in 
August 1968 were recaptured a few weeks later in the 
East; three birds were found beyond the Urals at a maxi- 
mum distance of almost 3,000 km (Zink 1981). Invaders 
stayed at wintering places in Bielefeld, Germany, when 
Siberian pine seeds (Pinus sibirica) were offered ad 
libidimi (Conrads and Balda 1979). 

Outside the natural range of cembra pine (Central Alps, 
Tatra, and a few locations in the Carpathian mountains) 
the European nutcracker normally feeds on hazel nuts 
iCorylus avellana). Slight, but statistically significant, 
differences in bill size are apparent in central-alpine 
populations. Nutcrackers from the Engadine and migrat- 
ing birds at Col de Bretolet in southwestern Switzerland 
show a narrower lower mandible and a longer bill com- 
pared with populations from southern Germany and from 
northern Europe (Mattes 1978; Mattes and Jenni 1984). 
Also, there are distinct differences in bill size in popula- 
tions of A^. c. macella from Szetchuan and from Sikkim 
(Mattes, unpublished). Generally, there seems to be a 
correlation of width of lower mandible and size of main 
food. The distance between the branches of the lower 
mandible limits the size of the entrance to the sublingual 
pouch. As the main food of most Asian populations of the 
nutcracker is not yet sufficiently known, bill size and bill 
structure cannot yet be fully explained. Furthermore, bill 
size in some subspecies may reflect feeding conditions at 
former times. 



REFERENCES 

Balda, Russel P.; Kamil, Alan C. 1989. A comparative 

study of cache recovery by three corvid species. Animal 

Behavior. 38: 486-495. 
Bibikov, D. I. 1948. On the ecology of the nutcracker. 

Trudy Pechoro-Ilychkoge Gosud. Zapovednik. 4: 89-112. 
Biu-ckhardt, D. 1958. Vom Vorratanlegen des Tannen- 

hahers. Biindnerwald. 11:102-114. 
Conrads, Klaus; Balda, Russell P. 1979. Uberwinter- 

imgschancen Sibirischer Tannenhaher {Nucifraga 

caryocatactes macrorhynchos) im Invasionsgebiet. 

Bericht des Naturwissenschaftlichen Vereins Bielefeld. 

24: 115-137. 

Crocq, Claude. 1990. Le casse-noix mouchete {Nucifraga 
caryocatactes). Paris: Lechevalier/R. Chabaud. 326 p. 

Grabher-Meyer, Arno. 1991. Beobachtungen zur Brut- 
biologie einer kleinen Population des Tannenhahers 
{Nucifraga caryocatactes) bei Praxmar, Sellrain. 
Diplomarbeit, Universitat Innsbruck, Austria. 

Kamil, Alan C; Balda, Russell P. 1988. Spatial memory 
in Clark's nutcracker {Nucifraga columbiana). Acta XIX 
Congressus Internat. Ornithol., Ottawa 1986. Vol. II: 
2100-2111. 

Lanner, Ronald M. 1982. Adaptations of whitebark pine 
for seed dispersal by Clark's nutcracker. Canadian 
Journal of Forest Research. 12(2): 391-402. 

Marzluff, John M.; Balda, Russ P. 1992. The pinyon jay: 
behavioral ecology of a colonial and cooperative corvid. 
London: Poyser. 317 p. 

Mattes, Hermann R. 1978. Der Tannenhaher im Engadin. 
Studien zu seiner Okologie und Funktion im Arven- 
wald. Miinstersche Geographische Arbeiten 2. 
Paderbom: Schoning. 87 p. 

Mattes, Hermann R.; Jenni, Lukas. 1984. Ortstreue und 
Zugbewegungen des Tannenhahers {Nucifraga caryo- 
catactes) im Alpenraima und am Randecker Maar/ 
Schwabische Alb. Ornithologischer Beobachter. 81: 
303-315. 

Oberauer, Eva. 1991. Zur Postembryonalentwicklimg des 
Tannenhahers {Nucifraga caryocatactes L.). Beobach- 
tungen im Alpenzoo Innsbruck. Diplomarbeit, Uni- 
versitat Innsbruck, Austria. 140 p. 

Reijmers, N. F. 1959. Birds of the cedar-pine forests of 
south-central Siberia and their role in the life of the 
cedar-pine. Trudy biolog. Inst, sibirsk. Otdel. Akad. 
Nauk SSSR. 5: 121-166. 

Rudat, Volker; Rudat, Wolf. 1978. Zum Verhalten von 
Tannenhaherfamilien {Nucifraga caryocatactes L.) in 
der Zeit vom Ausfliegen der Jungen bis zu deren 
Selbstandigwerden. Zoologisches Jahrbuch fur 
Systematik. 105: 386-398. 

Saito, Shin-ichiro. 1983. On the relations of the caching 
by animals on the seed germination of Japanese stone 
pine, Pinus pumila Regel. Bulletin of the Shiretoko 
Museum. 5: 23-40. 

Swanberg, Peer O. 1951. Food storage, territory and song 
in the thickbilled nutcracker. Acta X Congressus 
Internat. Ornitholog. Uppsala 1950: 545-554. 

Swanberg, Peer O. 1956. Territory in the thickbilled nut- 
cracker {Nucifraga caryocatactes). Ibis. 98: 412-419. 



34 



Tomback, Diana F. 1977. Foraging strategies of Clark's 

nutcracker. The Living Bird. 17: 123-161. 
Tomback, Diana F. 1982. How nutcrackers find their seed 

stores. The Condor. 82: 10-19. 
Tomback, Diana F.; Linhart, Yan B. 1988. The evolution 

of bird-dispersed pines. Evolutionary Ecology. 4: 

185-219. 

Vander Wall, Stephen B. 1982. An experimental analysis 
of cache recovery in Clark's nutcracker. Animal 
Behavior. 30: 84-94. 

Vander Wall, Stephen B.; Balda, Russ P. 1977. Coadapta- 
tions of the Clark's nutcracker and the pinyon pine for 



efficient seed har\'est and dispersal. Ecological Mono- 
graphs. 47(1): 27-37. 
Vander Wall, Stephen B.; Balda, Russ P. 1981. Ecology 
and evolution of food storage behavior in conifer- 
seed-caching Corvids. Zeitschrift fiir Tierpsychologie. 
56: 217-242. 

Woolfenden, D. E.; Fitzpatrick, W.-J. 1984, The Florida 
scrub jay. Princeton, NJ: Princeton University Press. 

Zink, Gerhard. 1981. Der Zug europaischer Zugvogel. 3. 
Liefenmg. Moggingen: Vogelzug-Verlag. 



35 



ALLOZYME POLYMORPHISM, 
HETEROZYGOSITY, AND MATING 
SYSTEM OF STONE PINES 

Dmitri V. Politov 
Konstantin V. Krutovskii 



Abstract— Population genetic structure and the mating system of 
stone pines (genus Pinus, subsection Cembrae) have been studied 
using isozymes. Observed were a tendency to heterozygote defi- 
ciency among embryos (obviously, a consequence of self-pollination 
with the estimated fi-equency of 0.11-0.30) and a slight heterozy- 
gote excess in adult trees (probably caused by selection against 
inbred progeny and balanced selection). 



Electrophoretic analysis of allozjone polymorphism is 
one of the most useful methods for study of genetic and evo- 
lutionary processes in natural populations and taxonomic 
and phylogenetic relationships of forest tree species (Adams 
and others 1992; Fineschi and others 1991; MuUer-Starck 
and Ziehe 1991). Conifers are particularly attractive, from 
a genetic point of view, because of their haploid megaga- 
metophyte (endosperm). Electrophoresis of both haploid 
and diploid (embryo) seed tissues permits us to distinguish 
maternal and paternal contributions into the progeny geno- 
type. This makes it possible to estimate allele frequencies 
separately in both male and female gametes pools and to 
study segregating ratios directly by analyzing endosperm 
haplotypes, which correspond to haplotypes of female ga- 
metes. The possibility of observing meiotic products di- 
rectly through study of haploid endosperms simplifies analy- 
sis of allozyme inheritance, genetic Unkage of isozyme loci, 
and the mating system of conifers. 

In contrast to many widespread European and North 
American conifer species, stone pines have been studied 
using isozymes only recently (Groncharenko and others 1988; 
Krutovskii and Politov 1992; Krutovskii and others 1987, 
1988, 1989, 1990; Podogas and others 1991; Politov 1989; 
PoUtov and Krutovskii 1990; Politov and others 1989, 1992; 
Szmidt 1982). Meanwhile, the need for such data is appar- 
ent both for practical forestry applications and for purposes 
of genetic diversity conservation. As a major forest-forming 
tree species, stone pines determine, to a large extent, the 
state of ecosystems over extensive areas. Their genetic vul- 
nerability thus threatens the existence of a unique forest 
type. 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Dmitri V. Politov is Research Scientist and Konstantin V. Krutovskii 
is Senior Research Scientist, Laboratory of Population Genetics, Russian 
Academy of Sciences, Gubkin Str. 3, 117809 GSP-1 Moscow, Russia. 



Four of the five stone pine species traditionally united 
in subsection Cembrae of genus Pinus (Critchfield and 
Little 1966) are found in Russia and neighboring countries: 
Siberian stone pine (Pinus sibirica Du Tour), European or 
Swiss stone pine (P. cembra L.), Korean pine {P. koraiensis 
Siebold et Zucc), and mountain or Japanese stone pine 
(P. pumila [Pall.] Kegel). All of them are very important 
economically. 

This paper reports the main results of long-term popula- 
tion genetic studies on these species using isozyme analy- 
sis. Some population structure parameters also have been 
estimated for one population of whitebark pine (P. albicaulis 
Engelm.), the only representative of stone pines in the 
Western Hemisphere. 

MATERIALS AND METHODS 

Sampling Sites— Seeds collected in natural populations 
served as a source of laboratory material. Pinus sibirica 
seeds were collected from 11 localities. Samples desig- 
nated as Filin Klyuch (F), Mutnaya Rechka (M), Maljd 
Kebezh (MK), Sobach'ya Rechka (SR), and Listvyanka (L) 
originate from the Western Sayan (Ermakovskii District, 
Krasnoyarsk Territory); Yayl'u (YA) from the Altfd Moun- 
tains (Gomo-Altaiskaya Autonomous Region); Smokotino 
(S), Zorkal'tsevo-1 (Z-1), and Zorkal'tsevo-3 (Z-3) from Tom- 
skaya Region; Turukhan (TU) from the north of Krasnoy- 
arsk Territory (Turukhanskii District); and N'alino (NYA) 
from the Khanty-Mansiiskii Autonomous District (Tjoimen- 
skaya Region). Besides, one locality of P. cembra, Ust'- 
Chornaya (UCH) from the Eastern Carpathians (Ukraine); 
three localities of P. pumila from the north of Kamchatka 
Peninsula designated as Grustnyi (GR), Uttyveem (UT), 
and Kamenskoye (K); and three localities of P. koraiensis 
from Khabarovsk Territory designated Malokhekhtsirskoye 
(MKH) and Khor (KHO), and from Primorskii Territory 
(Sikhote-Alin' (SA)) have been also studied. The sample 
of P. albicaulis originates from the Livingstone Falls (LF) 
population (southwestern Alberta, Canada). 

Electrophoresis and Data Processing— No less than 
seven endosperms and embryos per tree have been studied 
using starch gel electrophoresis. Totally, 536 trees of five 
species (from 11 to 75 trees per population) and more than 
3,500 embryos (7 to 20 per tree), as well as two bulked sam- 
ples of seeds representing a large number of trees (TU and 
NYA), have been analyzed. Specimen preparation, buffer 
systems used, genetic interpretation of zymograms, and des- 
ignations of £dlozymes, alleles, and loci have been described 
elsewhere (Krutovskii and others 1987; Politov 1989). 



36 



Genotypes of particular trees were inferred from segre- 
gation of alleles among their endosperms. Then, allele and 
genot5T)e frequencies in samples of both mature trees and 
embryos, as well as in the effective pollen pool (paternal 
alleles of embryos), were estimated. Embryo tissue elec- 
trophoresis does not give high resolution for all zones of 
activity and, therefore, calculations based on embryo sam- 
ples have included a smaller nimiber of loci. Besides, some 
activity zones in individual species could not be reliably in- 
terpreted or were invariant. Because of this, specific sets 
of loci used for particular purposes in these species may dif- 
fer somewhat. More detailed description of statistic proce- 
dures can be fovmd elsewhere (Krutovskii and others 1989; 
Politov 1989; PoHtov and Krutovskii 1990; Politov and oth- 
ers 1989). 

GENETIC CONTROL OF ISOZYMES 
AND LINKAGE OF LOCI 

We have described inheritance of 11 isozyme systems 
of stone pines in our previous publications (Krutovskii and 
others 1987; Politov 1989) and foimd that they are controlled 
by at least 26 loci. However, only 21 most reliable loci with 
unambiguous genetic control have been included in this 
study (table 1). Segregation ratios of allelic variants of in- 
dividual allozjnne loci observed among endosperms of het- 
erozygous trees correspond, with several exceptions, to the 
expected ones (1:1). 

Analysis of joint allelic segregation for pairwise loci com- 
binations in Pinus sibirica has revealed genetic linkage 
of loci Adh-1 and Lap-3 (with the mean recombination fi^e- 
quency 0.275), Adh-1 and Pgi-2 (0.301), and Fe-2 and Lap-2 
(0.309). These loci belong to two linkage groups: Lap-3 — 
Adh-1— Pgi-2 and Lap-2— Fe-2 (Politov and others 1989). 
Comparison of these data with those available in literature 
shows that stone pines, as well as other representatives of 
the family Pi/iacme, are characterized by extreme conser- 
vatism of their genome, since the same loci combinations 
with nearly the same values of recombination fi-equency 
have been described not only in the closely related P. albi- 
caulis (Furnier and others 1986), but also in many hard 
pines (see, for example, Szmidt and Muona 1989). Data 
presented here are apparently the first attempt of genetic 
mapping in one of the stone pines. However, one can as- 
sume that the linkage groups revealed in P. sibirica will 
also be fotmd in other related species. Information of this 
kind could be useful for breeding programs, for analysis 
of association of genetic markers with economically impor- 
tant characters, and for mating system parameters estima- 
tion by multilocus methods. 

POPULATION GENETIC STRUCTURE 

Levels of Intrapopulation Variability — ^According 
to the gene diversity parameters estimated in stone pine 
populations using isozjone loci, the level of genetic vari- 
ability appears to be rather high (table 2). There are al- 
most no differences between observed heterozygosities and 
those expected fi'om corresponding Hardy- Weinberg pro- 
portions. The values of the latter for individual species 
are: for P. sibirica— from 0.140 to 0.176, P. cembra— 0.109, 



Table 1 — Enzymes and loci analyzed 

Total 
number 



Enzyme (abbreviation) 


of loci 


Scored loci 


Alcohol dehydrogenase (ADH) 


2 


Adh-1 , Adh-2 


DIaphorase (DIA) 


3 


Dia-1 , Dia-2 


Fluorescent esterase (FE) 


3 


Fe-2 


Glutamate dehydrogenase (GDH) 


1 


Gdh 


G 1 utam atoxaloacetate 


3 


Got-1 , Got-2, Got-3 


transaminase (GOT) 






Isocitrate dehydrogenase (IDH) 


1 


Idh 


Leucine aminopeptidase (LAP) 


3 


LaD-2 LaD-3 


Malate dehydrogenase (MDH) 


4 


Mdh-1, Mdh-2, Mdh-3 






Mdh-4 


Phosphoglucose isomerase (PGI) 


2 


Pgi-2 


Phosphoglucomutase (PGM) 


2 


Pgm-1, Pgm-2 


Shikimate dehydrogenase (SKDH) 


2 


Skdh-1 , Skdh-2 



P. pumila — from 0.239 to 0.258, P. koraiensis — from 0.113 
to 0.148, P. albicaulis— 0.204. 

Such a high amoimt of intrapopulation variability is char- 
acteristic of conifers. We have summarized data on allozyme 
diversity of 30 pine species using available references 
(Politov and others 1992). The mean number of alleles per 
locus (A) and the expected heterozygosity (Hg), averaged for 
all pine species, equalled 2.08 and 0.169, correspondingly. 
Comparable values have been found for spruces (A = 2.2, 
Hg = 0.183), larches (A = 1.6, H^ = 0.091), and representa- 
tives of other less-studied genera. 

Factors supposedly promoting a high level of genetic diver- 
sity in stone pine populations include such ones, common 
for forest trees, as predominantly outcrossing mating sys- 
tem, wide distribution, large effective size of population, 
and high longevity (Hamrick and Godt 1989; Ledig 1986). 
More specific evolutionary factors, such as drastic reduc- 
tion of population size ("bottleneck effect"), may also affect 
the variability level. Although the ranges of stone pines, 
both in Eurasia and in North America, were reduced diu-- 
ing the Quaternary glacials (and also during the global 
warming), preservation of extensive refugiimis permitted 
avoidance of substantial genetic depauperization. This 
concerns Pinus cembra to a much less degree than other 
stone pines. 

According to some evidence, Swiss stone pine imderwent 
a sharp reduction in its range and population size during the 
Holocene caused by the global warming (Bobrov 1978 and 
references therein). A small effective number is typical of 
most populations of this species at present as well. Swiss 
stone pine exists nowadays in the Alps and the Carpathians 
as small isolated stands in which processes of genetic drift 
do not allow aUozyine variability to be maintained on a high 
level. This is a most probable cause for a lower level of het- 
erozygosity (0.109) and the percentage of polymorphic loci 
(26.3 percent) of P. cembra, as compared to other species. 
However, it should be noted, that the only sample of this spe- 
cies studied by us represents a small marginal stand from 
the eastern Carpathians, while central populations may 
appear to be more variable. The limited nimaber of loci 
studied in the only work (Szmidt 1982) dealing with allo- 
zyme polymorphism in Swiss stone pine over the species 



37 



Table 2 — Levels of allozyme variability in stone pine populations 



Species and Number Number 



population names 


of loci 


of trees 




P2 


Ho' 


He' 






Dim 1^ ^ihifS^o 
ninus SiOiriCa 










Filin Klyuch^ (F«) 


19 


53.0 


1.8 


47.4 


0.155 


0.152 


Mutnaya Rechka (M) 


19 


75.0 


1.9 


47.4 


.137 


.140 


Maiyi Kebezh (MK) 


19 


15.0 


1.8 


52.6 


.158 


.153 


Sobach'ya Rechka (SR) 


19 


12.0 


1.7 


42.1 


.145 


.173 


Listvyanka (L) 


19 


41.0 


1.9 


42.1 


.163 


.151 


Yayl'u (YA) 


19 


43.0 


1.9 


47.4 


.153 


.154 


Smokotino (S) 


19 


43.0 


1.7 


42.1 


.168 


.161 


Zorkal'tsevo-I (Z-1) 


19 


34.0 


1.7 


47.4 


.198 


.176 


Zorkartsevo-3 (Z-3) 


19 


34.0 


1.7 


47.4 


.184 


.165 


Mean 


19 


38.9 


1.8 


46.2 


.162 


.158 
















UsfChornaya (UCH) 


19 


15.0 


1.5 


26.3 


.128 


.109 






Pinus pumila 










Grustnyi (GR) 


20 


54.0 


2.3 


50.0 


.230 


.239 


Uttyveem (UT) 


20 


55.0 


2.5 


60.0 


.243 


.251 


Kamenskoye (K) 


20 


55.0 


2.3 


70.0 


.269 


.258 


Mean 


20 


54.7 


2.4 


60.0 


.247 


.249 






Pinus koraiensis 










Malokhekhtsirskoye (MKH) 


17 


30.0 


1.6 


47.1 


.147 


.132 


Khor (KHO) 


17 


16.0 


1.6 


47.1 


.162 


.148 


OlKllOie-Mlin ^OA; 


1 7 
1 / 


1 1 .u 


1 .O 






.Mo 


Mean 


17 


19.0 


1.6 


43.2 


.137 


.131 






Pinus albicaulis 










Livingstone Falls (LF) 


16 


52.0 


1.9 


56.3 


.213 


.204 



'Mean number of alleles per locus. 

^Mean percent of loci polymorphic (locus has been considered polymorphic if the frequency of the most com- 
mon allele has not exceeded 95 percent). 
^Mean heterozygosity observed. 
"Mean heterozygosity expected. 
^See text for origins of populations. 
^Abbreviated denomination of population. 



range does not allow any conclusions to be made about the 
variability level. 

Heterozygosity of Korean pine (mean 0.131) is slightly 
higher. As to Pinus pumila, this species shows the highest 
values of H among stone pines (0.249), close to the maxi- 
mum ones known for conifers (Politov and others 1992). 
One cannot exclude the fact that such a high variability 
is very important for adaptation of this species to a wide 
spectnmti of environmental conditions, including those of 
permafrost regions, and drastic seasonal climatic changes. 
Nearly the same heterozygosity level (0.204) was found in 
the only studied population of whitebark pine from the 
north moimtain area. 

Genotype Distribution in Adult Trees and 
Embryos — Comparison of observed genotype distributions 
with those expected from the corresponding Hardy- Weinberg 
proportions is commonly used in population genetics to re- 
veal and identify various factors affecting genetic structure 
of populations. In stone pines such analysis has not re- 
vealed pronounced deviations from Hardy- Weinberg equi- 
librium for mature trees in any samples. 



Statistically significant excess or deficiency of heterozy- 
gotes have not been observed for allozyme loci studied in 
stone pine populations. Nevertheless, the sign criterion 
showed that there were significantly more cases of a slight 
heterozygote excess (Politov and others 1992), which may 
mean existence of a tendency toward systematic excess of 
heterozygotes of alloz3mie loci among adult trees. Figure 1 
shows mean values of Wright's fixation index (or "inbreed- 
ing coefficient," F = l-H/Hg, where H^ and Hg are mean 
observed and expected heterozygosities for all loci studied 
in the species). Predominance of slightly negative or almost 
zero F values indicates absence of significant heterozygote 
deficiency in the reproductive part of the popvilations. Spe- 
cies mean values were: for P. sibirica, -0.025; for P. cembra, 
-0.174; for P. koraiensis, -0.046. F values for P. pumila 
were slightly positive in two samples and negative in the 
third one (mean F = 0.008). 

In most embryo samples, on the contrary, a slight defi- 
ciency of heterozygotes was regularly observed (Politov and 
others 1992). Figure 2 presents mean F values in embryos 
and adult trees calculated for the same loci. In embryos 



38 



0.1 
0.05 

F 0 

-0.05 
-0.1 
-0.15 







A. 
'1 


niri 


1 


ij 








1 

















Populations: F M MK SR L YA S Z-1 Z-3 MEAN 



GR UT K MEAN MKH SA MEAN 



J KHO 
I 



J 



P. koraiensis 



P. sibirica P. cembra P. pumila 

Figure 1 — Observed fixation index ("inbreeding co- 
efficient") values (F) in samples of adult trees. All 
studied loci used. (For full names and origins of 
populations see text and table 2). 



the values are positive in nearly all samples (except one), 
whereas mature trees display mainly negative F values. 

Probable Causes of Heterozygote Deficiency — 

most probable cause for heterozygote deficiency at early de- 
velopmental stages of stone pines may be inbreeding. The 
species studied, as well as other conifers, have a mixed mat- 
ing system (Brown and others 1985; Politov and Krutovskii 
1990 and references therein) characterized by a random 
outcrossing with a certain percentage of selfing. Heterozy- 
gote deficiency may theoretically be caused by both self- 
pollination and other consanguineous matings (Ritland and 
El-Kassaby 1985; Shaw and others 1981). One cannot also 
exclude the action of natural selection against heterozy- 
gotes at a certain developmental stage. 

Much evidence has been reported demonstrating hetero- 
zygote deficiency among embryos in populations of conifers 
(summarized by Bush and Smouse, 1992). In some rare 
cases deficiency is revealed in mature trees. The authors 
explained this phenomenon by inbreeding and, particularly, 
by selfing. 

Mating System Analysis — ^The fact that selfing is most 
commonly regarded as a factor causing inbreeding in conifer 
species has good reasons. Coincidence, in general, of polli- 
nation with pistil receptivity of the same tree and absence of 
effective self-sterility mechanisms make this process highly 
probable. In order to test the h3T)othesis that partial selfing 
is the main cause of heterozygote deficiency in stone pines, 
we have applied methods of mating system parameters es- 
timation based on allozyme data and mixed mating system 
theory (Brown and others 1985). Use of isozymes and pow- 
erful mathematical procedures (Expectation-maximization 
method, Newton-Raphson method, bootstrap, etc.) have 
made it possible to obtain quantitative estimates of selfing 
and outcrossing rates in natural conifer populations. For 
different species percentage of selfing varies between 2 to 3 



and 20 to 30 percent (see Brown and others 1985; Politov 
and Krutovskii 1990, for review). For stone pines such 
estimates have not been made until recently. Here we 
present data obtained on three stone pine species. 

The results of mating system analysis of Siberian stone 
pine have been published earlier (Politov and Krutovskii 
1990). If we consider a randomly outcrossing population 
at eqmlibrium, absence of pre- and postzygotical selections, 
sexual symmetry, separated generations, etc., the expected 
inbreeding coefficient Fg can be derived from the equation 
Fg = (l-^)/(l+^), where t is the outcrossing rate estimated 
for this species. In our study the expected inbreeding coef- 
ficient Fg has turned out to be very close to the actually 
observed one (F^) [F„ = iK~R^)fH^, where Hg and H^ are, 
correspondingly, the expected from Hardy- Weinberg pro- 
portions and the directly observed heterozygosities] in em- 
bryo samples. Thus, selfing (1-t) whose estimated rate is, 
on the average, as high as 15 percent (from 2 to 27 in indi- 
vidual populations) is most likely to be one of the main fac- 
tors of heterozygote deficiency occurrence among embryos 
of this species. 

Available data for three stone pine species are presented 
in table 3. Estimation of outcrossing rates was made based 
on four to eight polymorphic loci and using two different 
computation procedures — variants of both single- and mtdti- 
locus methods (Ritland and El-Kassaby 1985; Ritland and 
Jain 1981). Both procedures have been realized by Ritland 
(1990) in his original computer program MLT. The multi- 
locus estimation (which has more detective power than the 
single-locus one) gives for Pinus sibirica populations values 
ranging from 0.846 to 0.980 (mean 0.894), and for P. korai- 
ensis values fi:'om 0.920 to 1.034 (mean 0.974). The out- 
crossing rate foimd in P. cembra was substantially lower 




P. sibirica 



P. cembra i_ 



Q Adult trees 
I Embryos 



Figure 2 — Observed fixation index, "inbreeding 
coefficient" (F^), values in samples of adult trees 
(striped bars) and embryos (filled bars). Only 
eight common loci used. Fg is expected inbreed- 
ing coefficient [Fg = {^-tJ^{^+tJ]. S is selfing 
rate (s = l-f^, see table 3). (For full names and 
origins of populations see text and table 2). 



P. lioraiensis 



39 



Table 3 — Multilocus (f J and single-locus {Q estimates of outcrossing 
rate in stone pine populations 



Species and Number Outcrossing rate 

population names of loci t„ (s.e.) (s.e.) 





Pinus sibirica 








Filin Klyuch-85^ 


8.0 


0.817 


(0.031) 


0.788 


(0.030) 


Filin Klyuch-86 


8.0 


.892 


(.042) 


.884 


(.052) 


Mutnaya Rechka-86 


8.0 


.980 


(.019) 


.959 


(.048) 


iviuiiidyct nuoiirVd 0/ 


8.0 


.863 


\.\JOO) 


.0^0 


1 n'^fi^ 


Maiyi Kebezh 


8.0 


.960 


(.072) 


.933 


(.060) 


Sobach'ya Rechka 


8.0 


.846 


(.125) 


.834 


(.105) 


^lOlVyCII IrVCl 


8.0 


.912 


(.031) 


fl71 




Yayl'u 


8.0 


.855 


(.045) 


.776 


(.043) 


Smokotino 


7.0 


.929 


(.033) 


.886 


(.037) 


Mean 


7.9 


.894 


(.057) 


.862 


(.054) 




Pinus cembra 








Ust'-Chornaya 


4.0 


.686 


(.025) 


.707 


(.045) 




Pinus 


koraiensis 






Malokhekhtsirskoye 


6.0 


.967 


(.062) 


.914 


(.052) 


Khor 


7.0 


.920 


(.037) 


.929 


(.066) 


Sikhote-Alin' 


6.0 


1.034 


(.070) 


.964 


(.031) 


Mean 


6.3 


.974 


(.058) 


.936 


(.051) 



'Number means year of sampling of material; see text for origins of 
populations. 



(0.686), but we regard this value only as a preliminary one 
because of the limited sample size and number of loci stud- 
ied. Nevertheless, it may also be the real level of selfing, 
probably due to the limited nimiber of trees in the studied 
locality and a low stand density. 

Shaw and others (1981) note that we can conclude the role 
of inbreeding components other than selfing per se in con- 
sanguineous matings (for example, mating between closely 
related neighbor trees in a population with family structure 
or assortative/selective pollination/fertilization between 
genetically similar trees) from the difference between the 
values of multi- and single-locus estimates. In oiu* study of 
stone pines these differences, for nearly all samples, appesir 
to be negligible (the difference, in general, does not exceed 
0.04). This implies a rather weak effect of mating between 
relatives or, in other words, absence of microsubdivision of 
the population. 

Fumier and others (1987) found that genetic and physical 
distances between clumps of Pinus albicaulis fail to corre- 
late. The authors related this fact to the effective "mixing" 
of seeds in the process of their dispersal by the Clark's nut- 
cracker {Nucifraga columbiana Engelm). We can expect to 
observe randomization of such kind (more closely spaced 
seedlings or clumps are not more related ones) in any other 
pine species depending on the European nutcracker (AT. cary- 
ocatactes L.) in their dispersal, since the effectiveness of nut- 
crackers as seed dispersers has been repeatedly proved by 
numerous investigations (Lanner 1980, 1990; Tomback 
and Linhart 1991). 

Selection Against Inbred Progeny — What may cause 
"disappearance" of heterozygote deficiency observed among 
embryos, as compared to adult trees of the population 



(figs. 1 and 2)? This can, most likely, be the result of selec- 
tion against inbred progeny formed by self-pollination, as 
well as selection in favor of heterozygotes (overdominance). 
Both selections may increase heterozygosity up to the equi- 
librium level, but only as a result of negative assortative 
fertilization or selection in favor of heterozygotes (balancing 
selection) can heterozygosity also exceed Hardy-Weinberg 
expectations. This process can be detected with the aid of 
isozyme gene markers if they are either involved in it di- 
rectly or £ire in gametic linkage disequilibrium with selec- 
tively significant genes (Crow and Kimura 1970). 

In naturgJ conifer populations heterozygosity of adult 
trees is often higher than in embryos (see Bush and Smouse 
1992; Krutovskii and others 1988; Politov and others 1992, 
for review). For instance, in Pinus radiata (Plessas and 
Strauss 1986) heterozygosity of yoimg trees (17 to 20 years) 
exceeded that of 5-year seedlings and in the latter it was 
higher than in embryos. Yazdani and others (1985) found 
higher heterozygosity of young £ind adult Scots pine trees 
as compared to embryos; differences between the young and 
the adult trees were slight indicating the action of selec- 
tion exactly at early stages. For stone pines we can suggest 
a pEirticular life stage of natural selection acting against 
homozygotes — ^when groups of seedlings begin to germinate 
out of nutcrackers' caches competing for resources. It is 
noteworthy that Eurasian stone pines (as compared to white- 
bark pine) display multistem forms rarely, which may be 
the result of intensive competition between seedlings of the 
same cache. 

More direct evidence was obt£iined in experiments carried 
out by Farris and Mitton (1984), who studied embryos and 
seedlings of yellow pine {Pinus ponderosa) and foimd the 
increase of heterozygosity with age. Muona and others 
(1987) reveeded heterozygosity deficit in Scots pine embryos 
and its "disappearance" in 1.5-year-old seedlings. 

Inbred progeny elimination at early developmental stages 
in conifers is also supported by some nonallozyme data (see 
for example, Koski 1973). 

CONCLUSIONS 

Study of alloz5rme variability in stone pines has shown 
the genetic diversity tjrpical to widespread and abimdant 
conifer species with continuous ranges. Differences between 
stone pines can be explained by both the events of their 
recent evolutionary history and adaptive significance of 
the allozyme variation. Heterozygote deficiency among em- 
bryos and slight heterozygote excess in mature trees are 
common features of stone pines, as well as other conifers. 
A substantial part of self-pollinated progeny is eUminated 
during ontogenesis by natural selection. 

ACKNOWLEDGMENTS 

We are greatly thankful to G. R. Furnier, V. N. Vorobjev, 
N. A. Vorobjeva, L. I. Milyutin, R. T. Gut, and P. Raztmiov 
for providing seeds and aid in their sampling, to Yu. S. 
Belokon' for helpful assistance in the laboratory analysis, 
to Yu. P. Altukhov for constant support of our study and 
very useful conmients on this meinuscript, and to K. Ritland 
for kindly providing us with the computer program MLT. 



40 



We also appreciate H.-R. Gregorius's invaluable comments 
that helped to substantially improve the manuscript. 

REFERENCES 

Adams, W. T.; Strauss, S. H.; Copes, D. L.; Griffin, A. R., 
eds. 1992. Population genetics of forest trees: Proceed- 
ings of the International Symposiimi on Population Ge- 
netics of Forest Trees; 1990 July 31-August 2; Corvallis, 
OR, U.S.A. Kluwer Academic Publishers. 420 p. (see also 
New Forests. 1992. 6.) 

Bobrov, E. G. 1978. Forest-forming conifers of the USSR. 
Moscow: Nauka. 190 p. [In Russian]. 

Brown, A. H. D.; Barrett, S. C. H.; Moran, G. F. 1985. Mat- 
ing system estimation in forest trees: Models, methods 
and meanings. In: Gregorius, H.-R., ed. Population genet- 
ics in forestry. Lecture Notes in Biomathematics. Vol. 60. 
Berlin-Heidelberg-New York-Tokyo: Springer- Verlag: 
32-49. 

Bush, R. M.; Smouse, P. E. 1992. Evidence for the adaptive 
significance of allozymes in forest trees. New Forests. 6: 
179-196. 

Critchfield, W. B.; Little, E. L. 1966. Geographic distribution 
of the pines of the world. Misc. Publ. 991. Washington, 
DC: U.S. Department of Agriculture. 99 p. 

Crow, J. F.; Kimura, M. 1970. An introduction to popula- 
tion genetics theory. New York: Harper and Row. 

Farris, M. A.; Mitton, J. B. 1984. Population density, out- 
crossing rate and heterozygote superiority in ponderosa 
pine. Evolution. 38(5): 1151-1154. 

Fineschi, S.; Cannata, F.; Malvolti, M. E.; Hattemer, H. H., 
eds. 1991. Biochemical markers in the population genet- 
ics of forest trees: Proceedings of the Joint Meeting of the 
Working Parties S2. 04-021 Population Genetics and Eco- 
logical Genetics and S2. 04-05 Biochemical Genetics of 
the International Union of Forest Research Organizations 
(lUFRO); October 1988; Porano, Italy. SPB The Hague: 
Academic PubHshing bv. 251 p. 

Fumier, G. F.; Knowles, P.; Aleksiuk, M. A.; Dancik, B. P. 

1986. Inheritance and linkage of allozymes in seed tissue 
of whitebark pine. Canadian Journal of Genetics and 
Cytology. 28(4): 601-604. 

Fumier, G. R.; Knowles, P.; Clyde, M. A.; Dancik, B. P. 

1987. Effect of avian seed dispersal on the genetic struc- 
ture of whitebark pine popidations. Evolution. 41(3): 
607-612. 

Goncharenko, G. G.; Padutov, V. E.; Krutovskii, K V.; 
Podzharova, Z. S.; Kirgizov, N. Ya.; Politov, D. V. 1988. 
Level of genetic variability in Pinus sibirica in Altai. 
Doklady Akademii Nauk SSSR. 299(1): 222-225. [In 
Russian]. [English translation published by Pleniim 
PubHshing Corp. in Doklady Biological Sciences; 1988 
September: 139-141]. 

Hamrick, J. L.; Godt, M. J. 1989. Allozyme diversity in 
plant species. In: Brown, A. H. D.; Clegg, M. T.; Kahler, 
A. L.; Weir, B. S., eds. Plant population genetics breeding 
and genetic resources. Sanderland, MA: Sinauer Associ- 
ates: 43-63. 

Koski, V. 1973. On self-fertilization, genetic load and sub- 
sequent inbreeding in some conifers. Comm. Inst. For, 
Fenn. 78(10): 1-40. 



Krutovskii, K. V.; Politov, D. V.; Altukhov, Yu. P. 1987. 
Genetic variability in Siberian stone pine, Pinus sibirica 
Du Tour. I. (jenetic control of isozyme systems. Genetika. 
23(12): 2216-2228. [In Russian]. [EngHsh translation 
published by Plenimi Publishing Corp. in Soviet Genet- 
ics; 1988 June: 1568-1576]. 

Krutovskii, K V.; Politov, D. V.; Altukhov, Yu. P. 1988. 
Genetic variability in Siberian stone pine, Pinus sibirica 
Du Tour. II. Level of allozyme variability in Western 
Sayan. Genetika. 24(1): 118-124. [In Russian]. [EngHsh 
translation published by Plenum Publishing Corp. in 
Soviet Genetics; 1988 July: 81-86]. 

Krutovskii, K V.; Politov, D. V.; Altukhov, Yu. P.; Milyutin, 
L. I.; Kxiznetsova, G. V.; Iroshnikov, A. I.; Vorob'ev, V. N.; 
Vorob'eva, N. A. 1989. Genetic variability in Siberian 
stone pine, Pinus sibirica Du Tour. IV. Genetic diversity 
and differentiation between populations, (jenetika. 25(11): 
2009-2032. [In Russian]. [EngHsh translation pubHshed 
by Plenimi Publishing Corp. in Soviet Cjenetics; 1990 May: 
1343-1362]. 

Krutovskn, K. V.; Politov, D. V.; Altukhov, Yu. P. 1990. In- 
terspecific genetic differentiation of Eurasian stone pines 
for isoenzyme loci. Genetika. 26(4): 694-707. [In Russian]. 
[English translation published by Plenum Publishing 
Corp. in Soviet Genetics; 1990 October: 440-450]. 

Krutovskii, K V.; Politov, D. V. 1992. Study of intra- and 
interspecific genetic differentiation between Eurasian 
cedar pines using isozyme loci and multidimensional 
analysis. In: Symposium on molecular mechanisms of 
genetic processes: proceedings. Moscow: Nauka Publish- 
ers: 87-96. [In Russian]. 

Lanner, R. M. 1980. Avian seed dispersal as a factor in the 
ecology and evolution of limber and whitebark pines. In: 
6th North American Forest Biology Workshop: proceed- 
ings; 1980 August 11-13. Edmonton, AB: University of 
Alberta: 15-47. 

Lanner, R. M. 1990. Biology, taxonomy, evolution, and 
geography of stone pines of the world. In: Schmidt, 
Wyman C; McDonald, Kathy J., comps. Proceedings — 
symposimn on whitebark pine ecosystems: ecology and 
management of a high-mountain resource. Gen. Tech. 
Rep. INT-270. U.S. Department of Agriculture, Forest 
Service, Intermoimtain Research Station: 14-24. 

Ledig, F. T. 1986. Heterozygosity, heterosis, and fitness 
in outbreeding plants. In: Soule, M. E., ed. Conservation 
biology: the science of scarcity and diversity: 77-104. 

Mvdler-Starck, G.; Ziehe, M., eds. 1991. Genetic variation 
in Eiiropean populations of forest trees: proceedings of 
the workshop on genetic variation in forest tree popula- 
tions in Europe; 1990; Gottingen, Germany. Frankfurt 
am Main: Sauerlander's Verlag. 271 p. 

Muona, O.; Yazdani, R.; Rudin, D. 1987. Genetic change 
between life stages in Pinus syluestris: allozjone varia- 
tion in seeds and planted seedlings. Silvae Genetica. 
36(1): 39-42. 

Plessas, M E.; Strauss, S. H. 1986. Allozyme differentiation 
among populations, stands, and cohorts in Monterey Pine. 
Canadian Journal of Forest Research. 16(6): 1155-1164. 

Podogas, A. v.; Shurkhal, A. V.; Semerikov, V. L.; Rakitskaya, 
T. A. 1991. Genetic variabiHty of needle tissue enzymes 
of Siberian cedar pine (Pinus sibirica Du Tour). Genetika. 



41 



27(4): 695-703. [In Russian]. [English translation pub- 
lished by Plenum Publishing Corp. in Soviet Genetics; 
1991 October]. 
Politov, D. V. 1989. Allozyme polymorphism, genetic dif- 
ferentiation, and mating system in Siberian stone pine 
Pinus sibirica Du Tour. IVIoscow: Institute of General 
Genetics, Russian Academy of Sciences. 190 p. Thesis. 
[In Russian]. 

Pohtov, D. v.; Krutovskii, K. V.; Altukhov, Yu. P. 1989. 
Genetic variability in Siberian stone pine, Pinus sibirica 
Du Tour. III. Linkage relationships among isozyme loci. 
Genetika. 25(9): 1606-1618. [In Russian]. [EngHsh trans- 
lation published by Plenum Publishing Corp. in Soviet 
Genetics; 1990 IVIarch: 1053-1063]. 

PoHtov, D. v.; Krutovskii, K. V. 1990. Genetic variability 
in Siberian stone pine, Pinus sibirica Du Toiu*. V. Analy- 
sis of mating system. Genetika. 26(11): 2002-2011. [In 
Russian]. [English translation published by Plenmn Pub- 
hshing Corp. in Soviet Genetics; 1991 May: 1309-1316]. 

Pohtov, D. v.; Krutovskii, K. V.; Altukhov, Yu. P. 1992. 
Isozyme loci characteristics of gene banks of populations 
of cedar pines. Genetika. 28(1): 93-114. [In Russian]. 
[English translation published by Plenvun Publishing 
Corp. in Soviet Genetics; 1992 July: 76-95]. 

Ritland, K. 1990. A series of FORTRAN computer programs 
for estimating plant mating system. Journal of Heredity. 
81: 235-237. 



Ritland, K.; El-Kassaby, Y. A. 1985. The nature of inbreed- 
ing in a seed orchard of Douglas fir as shown by an effi- 
cient multilocus model. Theoretical and Applied Genetics. 
71: 375-384. 

Ritland, K.; Jain, S. 1981. A model for the estimation of 
outcrossing rate and gene frequencies using n indepen- 
dent loci. Heredity. 47(1): 35-52. 

Shaw, D. v.; Kahler, A. L.; Allard, R. W. 1981. A multilocus 
estimator of mating system parameters in plant popula- 
tions. Proceedings of the National Academy of Sciences 
(USA). 78: 1298-1302. 

Strauss, S. H.; Bosquet, J.; Hipkins, V. D.; Hong, Y.-P. 1992. 
Biochemical and molecular genetic markers in biosys- 
tematic studies of forest trees. New Forests. 6: 125-158. 

Szmidt, A. E. 1982. Genetic variation in isolated popula- 
tions of stone pine {Pinus cembra). Silva Fennica. 16(2): 
196-200. 

Szmidt, A. E.; IMuona, O. 1989. Linkage relationships of 
allozyme loci in Pinus sylvestris. Hereditas. Ill: 91-97. 

Tomback, D. F.; Linhart, Y. B. 1991. The evolution of bird- 
dispersed pines. Evolutionary Biology. 4: 185-219. 

Yazdani, R.; Lindgren, D.; Rudin, D. 1985. Gene dispersion 
and selfing frequency in a seed-tree stand of Pinus sylves- 
tris (L.). In: Gregorius, H. R., ed. Population genetics in 
forestry. Lecture Notes in Biomathematics. Vol. 60. Berlin- 
Heidelberg-New York-Tokyo: Springer- Verlag: 139-154. 



42 



GENETIC POPULATION STRUCTURE 
AND GROWTH FORM DISTRIBUTION 
IN BIRD-DISPERSED PINES 

Diana F. Tomback 
WiUiam S. F. Schuster 



Abstract — ^At least nine species of the 20 wingless-seed pines 
are dependent on the nutcrackers {Nucifraga) for seed dispersal. 
Seed dispersal by nutcrackers influences several aspects of popu- 
lation structure: (1) tree dispersion patterns and growth form dis- 
tribution, (2) genetic relationships among trees within popula- 
tions, and (3) genetic diversity among populations. Studies to 
date of three species, whitebark pine {Pinus albicaulis), limber 
pine {Pinus flexilis), and Swiss stone pine (Pinus cembra), pro- 
vide examples of intrapopulation genetic relationships and growth 
form distribution; other recent work provides some information 
for interpopulation comparisons. 



of parent trees, resulting in a haphazard distribution of 
seeds with respect to seed source (Tomback 1978, 1982). 
Also, nutcrackers often bury more than one seed per cache 
(mean of three or fotir seeds per cache for Clark's nut- 
cracker, Tomback and Linhart 1990), and the seeds may 
come from the same parent tree (Tomback 1988; Tomback 
and Knowles 1989). 

In contrast, most wind-dispersed seeds fall within about 
120 m of parent trees and are dependent on prevailing 
winds for dispersal (see, for example, McCaughey and oth- 
ers 1986). Updrafts and wind patterns associated with 
storms may send wind-dependent seeds over longer dis- 
tances and in other directions, and seed-storing rodents 
may move seeds aroimd locally (Vander Wall 1992). How- 
ever, because the seeds of these species usually move rela- 
tively short distances and establishment is often in small, 
disturbed patches, a family structure (local aggregations 
of related genotypes) frequently results within popula- 
tions (see, for example, Knowles 1984; Linhart 1989; 
Linhart and others 1981; see also Fumier and others 1987; 
Tomback and Linhart 1990). 

These different seed-dispersal modes, bird vs. wind, 
may lead to different genetic population structures. Ge- 
netic population structure resiilts from nonrandom spatial 
distribution of genotypes (Epperson 1990, and references 
therein). Plant populations may show structure on differ- 
ent scales, such as among populations and within popula- 
tions. The common methods for investigating genetic 
population structure are based on allozyme analysis, 
comparing populations or subpopxilations by means of ei- 
ther allele frequencies, Wright's or Nei's Gg.^, statistics, 
or spatial autocorrelation. 

Seed dispersal by nutcrackers influences several levels 
of popiilation structure in wingless-seed pines. Within 
populations, it affects both (1) tree dispersion patterns 
(whether individuals are clumped vs. spaced) and, conse- 
quently, growth form distribution and (2) genetic relation- 
ships among trees. Among populations it affects genetic 
diversity. In this paper, we explore the specific impacts 
of nutcracker seed-dispersal behavior on each of these 
aspects of population structure. 

GROWTH FORM DISTRffiUTION 



Of the 20 species of pines with large, wingless seeds 
(collectively considered "stone pines"), nine are known 
dependents on the nutcrackers, Nucifraga (Corvidae), for 
seed dispersal (Tomback and Linhart 1990 and references 
therein). Five of these bird-pines are in the Subsection 
Cembrae, two in Strobi, and two in Parry a; for classifica- 
tion see Critchfield and Little (1966). For several species 
of these bird-dependent pines, other corvids, squirrels, or 
other rodents may also make varying contributions to tree 
estabHshment (see, for example, Hayashida 1989; Ligon 
1978; Vander Wall 1992), but the factor in common to all 
dispersers is the tendency to remove seeds from the vicin- 
ity of parent trees and "scatterhoard" them (biuy small 
clusters of seeds in subterranean caches for future use). 
Clark's nutcrackers (A^. columbiana) bury one to 15 or 
more seeds per cache at distances of several meters to 
22 km from parent trees (Hutchins and Lanner 1982; 
Tomback 1978, 1982; Vander Wall and Balda 1977). Eur- 
asian nutcrackers (A^. caryocatactes) bury up to 24 seeds 
per cache and travel similar distances (see, for example, 
Bibikov 1948; Mattes 1982; Mezhenny 1961; Turcek 
1966). 

Thus, nutcrackers create a highly impredictable seed 
shadow, with seedhng establishment oft^n occurring sev- 
eral kilometers from parent trees (Tomback and others 
1990). Nutcracker-dispersed seeds may end up in small, 
isolated areas, particularly following disturbances such 
as fires. In large, open or disturbed areas, different nut- 
crackers may cache seeds from several difi"erent stands 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Diana F. Tomback is Associate Professor of Ecology, Department of 
Biology and Center for Environmental Sciences, University of Colorado 
at Denver, Denver, CO, 80217-3364. William S. F. Schuster is Forest Di- 
rector, Black Rock Forest Consortium, Box 483, Cornwall, NY 12518. 



Research to date on three species of bird-dispersed 
pines — whitebark pine (Pinus albicaulis) (Fumier and 
others 1987; Linhart and Tomback 1985), limber pine 
(P. flexilis) (Carsey and Tomback 1992; Linhart and 
Tomback 1985; Schuster and Mitton 1991), and Swiss 
stone pine (P. cembra) (Tomback and others 1993) — 



43 



indicates that trees within populations occur in three 
different growth forms. The single-trunk growth form 
is typical of conifers and consists of a single genet (indi- 
vidual). Also composed of a single genet, the multi-trunk 
growth form has two or more trunks either contiguous 
at the base or fused at the base or part way up the bole. 
Multi-trunk trees may be the consequence of a tendency 
toward side-branching and lack of apical dominance, par- 
ticularly under stressful environmental conditions, such 
as unstable soils, aridity, or mechanical damage to the 
leader (Schuster and Mitton 1991; Weaver and Jacobs 
1990). Although the tree cluster growth form is also 
composed of two or more contiguous or fused trunks, the 
trunks consist of different genets (individuals). This 
growth form results from the germination and survival 
of more than one seedling from a seed cache (growth form 
terminology from Tomback and others 1990). 

The latter two growth forms have similar morphologies 
and thus cannot be distinguished with certainty without 
genetic analysis (Schuster and Mitton 1991; Tomback and 
others 1993) (figs. 1, 2). Consequently, in field surveys of 
frequency distribution of different growth forms, we com- 
bined the multi-trunk and tree cluster forms together in 




Figure 1 — Swiss stone pine {Pinus cembra) tree 
cluster. Protein electrophoresis indicated that each 
trunk is a different genet. 




Figure 2 — Swiss stone pine {Pinus cembra) 
multi-trunk tree. Protein electrophoresis indi- 
cated that both trunks had the same genotype. 



the category "tree clumps" (term from Lanner 1980; the 
same as "multi-trunk trees of unknown origin," Tomback 
and others 1990). For reporting the results of field sur- 
veys of tree clump frequency in populations, we use the 
term "tree site" to refer to the small area of ground sup- 
porting either a single-trunk tree, a multi-trimk tree, or 
a tree cluster. 

Overall, tree clumps accounted for 19 to 90 percent of 
the tree sites surveyed in three species of bird-dispersed 
pines (whitebark, limber, and Swiss stone) (table 1). 
Whitebark pine in the two stands surveyed had the high- 
est percentage of tree sites occupied by clumps. In the 
two studies of limber pine combined, the percentage of 
tree sites with clvmips ranged from 24 to 48; the latter 
population occurred on a steep, southeast-facing slope 
with gravelly soil at 3,310 m elevation. Carsey and 
Tomback (1992) also examined 17 tree sites in a stand 
of krummholz limber pine above timberline (3,460 m eleva- 
tion) and found clumps at all tree sites. For Swiss stone 
pine, clumps were found at 19 to 30 percent of the tree 
sites and up to 45 percent of seedling sites. In the Piz da 
Staz study area, the highest percentage of Swiss stone 
pine clumps occurred at higher elevations, but this did 
not hold for the Muottas Muragl study area. 



44 



Table 1 — Percentage of tree clumps (combined multi-trunk trees and tree clusters) in surveyed populations of three species 
of bird-dispersed pines 



Species 






Maximum 


Percent 




and 




Tree 


trunks per 


sites 




location 


Elevation 


sites 


clump 


with clumps 


References 




Meters 










Swiss stone pine 












Switzerland 


1,835 


406 


6 


21 


Tomback and 




2,070 


351 


6 


30 


others 1 993 




1 ,835 


96 


4 


29 






2,125 


43 


5 


19 




Whitebark pine 












California 


2,850 


50 


5+ 


90 


Tomback and 


Montana 


2,730 


50 


5+ 


58 


Linhart 1990 


Limber pine 












Colorado 


1,650 


361 


7 


30 


Schuster and 












Mitton 1991 


Colorado 


2,585 


105 


3 


35 


Carsey and 




2,810 


135 


4 


24 


Tomback 1992 




3,310 


140 


4 


48 





The preferred laboratory method for distinguishing tree stone pine clumps sampled in one stand, 70 percent were 
clusters from multi-tnmk trees has been starch gel pro- tree clusters (Tomback and others 1993). Remarkably, 

tein electrophoresis to assay the genotype of seed (Fiimier an examination of the number of gene loci used for each 
and others 1987) or foliage tissue obtained from each study in table 2 shows no relationship with the percent- 

tnmk of a tree clump (Carsey and Tomback 1992; Linhart age of clusters determined for each sample (Spearman 
and Tomback 1985; Schuster and Mitton 1991; Tomback rank correlation, = -0.406, n = 12, NS). With the excep- 

and others 1993; for review of technique, see Kephart tion of the Umber pine clumps analyzed by Schuster and 

1990). Those clvunps that have two or more trimks with Mitton (1991), which represented all the climips in the 

different genotypes are classified as tree clusters. Unfor- popialation, the clvmips examined in the studies in table 2 

timately, this technique may underestimate slightly the 
nimibers of tree clusters, because seeds within nutcracker 
caches tend to be genetic relatives and thus may inherit 
similar genotypes (Carsey and Tomback 1992; Schuster 
and Mitton 1991). In a few cases tree cluster members 
may possess identical genotypes at the gene loci exam- 
ined, so the clusters are misclassified as multi-trunk 
(single-genotype) trees. The proced\ires for calculating 
the probability of such errors are presented in Schuster 
and Mitton (1991). 

For studies of whitebark pine, hmber pine, and Swiss 
stone pine combined, genetic analysis indicated that 18 
to 100 percent of the tree clumps sampled were tree clus- 
ters (table 2). The low value, 18 percent, was obtained by 
Schuster and Mitton (1991) for a limber pine population 
in the Pawnee National Grasslands, at the periphery of 
the range of both the pine and Clark's nutcracker. The 
high values of 100 percent for limber pine and 83 percent 
for whitebark pine were based on small sample sizes 
in = six clumps in both cases, Linhart and Tomback 1985) 
Filmier and others (1987) found that 58 and 70 percent, 
respectively, of the whitebark pine clumps they sampled 
in two stands contained more than one genotype; they did 
not sample all trunks of some clumps, so these percent- 
ages are conservative. Clumps in four limber pine stands 
were studied by Carsey and Tomback (1992), with per- 
centages of clusters ranging from 24 percent above tim- 
berhne to 81 percent at 2,810 m elevation. Of the Swiss 



Table 2 — Percentage of tree clumps (combined multi-trunk trees 
and tree clusters) found to be tree clusters (growth form 
with more than one genet) in three species of bird- 
dispersed pines 



Species 


Clumps 
examined 


Gene 
loci 
used 


Percent 
clusters 


References 


Swiss stone 


23 


4 


70 


Tomback and others 


pine 








1993 


Whitebark pine 


6 


4 


83 


Linhart and Tomback 










1985 




12 


11 


58 


Furnier and others 




23 


11 


70 


1987 


Limber pine 


7 


2 


57 


Linhart and Tomback 




6 


4 


83 


1985 




6 


4 


100 






108 


10 


18 


Schuster and Mitton 










1991 




18 


7-9 


56 


Carsey and Tomback 




21 


7-9 


81 


1992 




18 


7-9 


44 






17 


7-9 


24 





45 



represented a sample of the total number of climaps present 
and were sometimes not selected entirely at random. For 
the studies by Furnier and others (1987) and Carsey and 
Tomback (1992), only reproductively mature tree clumps 
were sampled; for the studies by Linhart and Tomback 
(1985) and Tomback and others (1993), clumps were usu- 
ally sampled as encountered, but an effort was made to 
include some seedlings and young trees. 

Only three studies permit an estimate of the overall 
percentage occurrence of tree clusters among tree sites 
in populations (table 3). Limber pine tree clusters in the 
Pawnee National Grasslands population occurred at only 
5 percent of the total tree sites. For limber pine in the 
Front Range, clusters occurred at similar percentages of 
tree sites, 19 to 21, despite differences in elevation. Clus- 
ters of Swiss stone pine occupied 21 percent of the tree 
sites in the upper subalpine population. 

Apparently, tree clusters typically occupy a substantial 
proportion of the tree sites in each population of whitebark, 
limber, and Swiss stone pine. Moreover, the percentage 
of individual genets in a population that are found in clus- 
ters is much higher than the proportion of tree sites sup- 
porting clusters. Based on the previous studies, 12 to 40 
percent of the individuals in populations of bird-dispersed 
pines occur in a highly clumped spatial pattern, primarily 
as a consequence of seed dispersal by nutcrackers. 

If stress plays a role in the growth of the multi-trunk 
form from a single genet, as proposed by Schuster and 
Mitton (1991) and Weaver and Jacobs (1990), then we 
might predict a relatively higher percentage of multi- 
trunk trees in more stressful environments. In fact, about 
25 percent of the limber pine tree sites in the Pawnee 
National Grasslands population support multi-trunk trees 
(from tables 1 and 2). This area is at lower elevations 
(1,650 m) on the plains east of the Rocky Mountains, at 
the eastern boundary of the limber pine range, and trees 
may be subjected to stress factors such as grazing, fire, 
unstable slopes, heat, and extreme aridity (Lepper 1974; 
McNaughton 1984; Schuster and Mitton 1991). The high- 
est estimated percentages of multi-trunk limber pine trees 
in the Front Range populations studied by Carsey and 
Tomback (1992) occurred in the upper subalpine popula- 
tion (27 percent, elevation 3,310 m) and in the population 
above timberline (76 percent, elevation 3,460 m), where 
winds and accompanying desiccation are extreme and the 



Table 3 — Estimated percentage of tree sites tliat support clusters of 
individuals in populations of two species of bird-dispersed 
pines 



Table 4 — Estimated mean and standard error of relatedness of 
individuals within and between limber pine tree clusters 



Percent 

Species clusters Elevation 



References 



Meters 

Swiss stone pine 21 2,050 to 2,240 Tomback and others 1 993 



Limber pine 



5 
19 
19 
21 



1,650 
2,585 
2,810 
3,310 



Schuster and Mitton 1 991 
Carsey and Tomback 1 992 



Mean Standard error 



References 



Within tree clusters 

0.19 0.10 Schuster and Mitton 1 991 

.43 .13 Carsey and Tomback 1 992 

Between tree clusters 

.01 .04 Carsey and Tomback 1 992 



soil is gravelly and unstable. We also calculated that only 
9 percent of the tree sites of the upper subalpine popula- 
tion of Swiss stone pine supported multi-trunk trees. In 
contrast to the former study areas, the Swiss stone pine 
tree sites are not typically characterized by aridity and 
poor soils, which may explain the relatively low frequency 
of multi-trunk trees (Tomback 1988). 

We should note here that tree clusters are not known 
to occur in Colorado pinyon pine (Pinus edulis), which is 
dispersed by both Clark's nutcrackers and pinyon jays 
(Gymnorhinus cyanocephalus). Vander Wall and Balda 
(1977) report that only one seed from each cache grows 
to maturity. 

GENETIC RELATIONSHIPS 

Relationships Within and Between Tree 
Clusters — Because nutcrackers tend to harvest and 
pouch several seeds from the same pine cone and visit 
more than one cone from the same tree or genet (if a clus- 
ter) (Tomback 1988; Tomback and Knowles 1989), they 
tend to place sibling seeds in the same seed cache. Dur- 
ing a bumper whitebark pine cone crop (1989) in Yellow- 
stone National Park, Tomback and Knowles (1989) ob- 
served nutcrackers taking entire pouchloads of 28 to 97 
seeds {x = 53 seeds, n = 8) from a single cone or individual 
tree (genet) or trimk. 

One consequence of this behavior is the tendency for 
individuals in tree clusters to be genetic relatives — either 
half-siblings, siblings, or the products of selflng. Furnier 
and others (1987) found that individuals within tree clus- 
ters of two populations were genetically more similar 
(smaller genetic distance) than individuals in different 
tree clusters. Individuals within clusters differed by a 
mean of 1.66 alleles; individuals of different clusters dif- 
fered by a mean of 5.82 alleles or greater (of 11 gene loci). 
Two studies examined genetic relatedness within limber 
pine clusters using the Quellar and Goodnight (1989) esti- 
mator (table 4): For tree clusters in the Pawnee National 
Grasslands population, Schuster and Mitton (1991) deter- 
mined that individuals were related on average as half- 
siblings. In addition, they determined that cluster mem- 
bers that had fused together at the base were more closely 
related on average (r = 0.35) than cluster members that 
had not fused (r = 0.05). Carsey and Tomback (1992) cal- 
culated that cluster members in their upper subalpine 
population (3,310 m elevation) were related on average 
as between half and full siblings; members of different 
clusters were unrelated on average (table 4). 



46 



Possible Biological Implications of Tree 
Clustering — Despite the fact that tree clusters occiir com- 
monly in populations of bird-dispersed pines and include 
a large proportion of the genets in the population, little is 
known of the biology of this growth form (Tomback and 
others 1993). Several papers have speculated about the 
potential advantages and disadvantages to individuals 
growing in clusters (Schuster and Mitton 1991; Tomback 
and Linhart 1990; Tomback and others 1993). 

Individuals in clusters may have higher survivorship 
than single genets vmder certain environmental condi- 
tions because of morphological or physiological benefits 
(Bullock 1981; Keeley 1988; Mattes 1982). A sturdy, well- 
anchored structure might better resist the strong winds 
and mechanical damage of higher elevation environments. 
Root grafting may also occur among cluster members, par- 
ticularly relatives (Schuster and Mitton 1991), resulting 
in efficient water and nutrient acquisition (Holtmeier 
1986). In addition, since nutcrackers may establish pines 
at some distance from conspecifics, for example, in burned 
areas or disturbed patches, individuals in tree clusters 
may cross-pollinate and thus avoid selfing (Tomback and 
Linhart 1990). This might produce a high proportion of 
sound seeds, particularly if clusters consisted of both rela- 
tives and nonrelatives. Finally, both Schuster and Mitton 
(1991) and Tomback and Linhart (1990) point out that the 
occurrence of kin groups of bird-dispersed pines may re- 
sult in kin selection as an evolutionary force. Possible 
consequences for cluster members may include some de- 
crease in competition or facilitation of rootgrafting. 

With respect to the potential disadvantages to indivi- 
duals growing in clusters, one area of possible fitness loss 
is reproductive output. For limber pine trees in three 
stands in the Colorado Front Range (same populations 
studied by Carsey and Tomback 1992), cluster members 
produced fewer male and female cones than did single- 
tnmk trees during each of 3 years (Feldman 1991; 
Feldman and Tomback 1991). Reduction of reproductive 
output may result from competition among cluster mem- 
bers for moisture, nutrients, and space. These effects may 
be exacerbated for limber pine, which typically grows on 
windy, xeric sites with poor soils in the Colorado Front 
Range (Feldman 1991; Feldman and Tomback 1991). 
Feldman (1991) and Feldman and Tomback (1991) noted 
that trees in clusters had smaller diameters than did 
similarly aged single-trunk trees, supporting the possibil- 
ity of competition. 

Another possible disadvantage to cluster members is 
that their seeds may be more inbred than seeds produced 
by single-genet trees, because of pollen exchange within 
clusters (Furnier and others 1987; Tomback and Linhart 
1990). PoUtov and Krutovskii (1990) provide some evi- 
dence for this in Siberian stone pine (Pinus sibirica), 
which is dispersed by the Eurasian nutcracker. Examin- 
ing the mating system of Siberian stone pine in seven 
populations, they noted an excess of homozygotes among 
seed embryos in the majority of their samples. They at- 
tributed this to a relatively high rate of self-pollination, 
but pollination among related cluster members would 
have a similar effect. 



Genetic Spatial Structure Within Popvilations — 

Several studies indicate that in many wind-dispersed 
conifers there is family structure within popiilations 
(local aggregations of genetic relatives) (see, for example, 
Knowles 1984; Linhart 1989; Linhart and others 1981). 
This is the consequence of seed dispersal typically re- 
stricted to short distances (Ftunier and others 1987; 
Tomback and Linhart 1990), and the availability of local, 
disturbed patches for colonization (see, for example, 
Linhart and others 1981). In contrast, Furnier and others 
(1987) did not find any evidence of family structure 
among the neighboring tree clusters in two whitebark 
pine populations (although cluster members tended to 
be related). They attributed this lack of family structiire 
with respect to geographic distance to the seed-caching 
behavior of Clark's nutcrackers. Since Clark's nutcrack- 
ers do not defend caching territories, several different 
nutcrackers may cache in one area, often bringing seeds 
in from different stands of trees (Tomback 1978, 1982). 
This probably results in a haphazard distribution of re- 
lated seed caches (Furnier and others 1987; Tomback 
and Linhart 1990). The Eurasian nutcracker in Sweden 
(Swanberg 1956) and in the Alps (Mattes 1982) stores 
many of its caches within territories defended by pairs. 
However, the seeds cached in these territories are both 
from local and more distant parent trees (up to 15 km. 
Mattes 1982), so there is also considerable mixing in any 
area. PoHtov and Krutovskii (1990) examined seven popu- 
lations of Siberian stone pine and also foimd no evidence 
for intrapopulation substructure. 

GENETIC DIVERSITY 

Generally speaking, conifers show less genetic diversity 
among popiilations (mean = 0.068) than do other kinds 
of plants (mean Gg^ = 0.250) (Hamrick and Godt 1990). 
After an extensive review of the plant literature, Hamrick 
and Godt (1990) concluded that animal-dispersed species 
have more differentiated populations (higher Gg^ values) 
than do wind-dispersed species. But, seed dispersal by 
nutcrackers, which occurs routinely over large distances, 
might actually result in higher levels of gene flow than 
from seed dispersal by wind and thus lower levels of popu- 
lation differentiation. 

Few studies have examined interpopulation variation in 
bird-dispersed pines. Schuster and others (1989) geneti- 
cally analyzed two limber pine populations at extreme 
ends of a 1,700-m elevational transect in the Colorado 
Front Range. They determined that pollen flow along this 
gradient was impeded by pollination phenology; popula- 
tions separated by more than 400 m elevation did not 
have overlapping pollination periods. For the populations 
at each end of the transect, allele frequencies differed sig- 
nificantly at eight of the 10 loci studied. However, the 
calculated nimiber of migrants exchanged between the 
populations per generation (A^^ = 11.1) was well above the 
level required to overcome genetic drift. Given the very 
different environmental conditions of the two populations, 
Schuster and others (1989) concluded that natural selec- 
tion rather than lack of gene flow may have differentiated 
the populations. They also suggested that seed dispersal 



47 



by Clark's nutcrackers is likely to be "a more important 
component of gene flow in limber pine than pollen trans- 
fer, at least across elevational gradients." 

Schuster and Mitton (1992) also compared five popula- 
tions of limber pine across a 100-km range and found low 
levels of differentiation (mean Gg^ = 0.035). Hamrick (un- 
published data cited in Betancourt and others 1991) ex- 
amined eight populations of pinyon pine (Pinus edulis) 
and found a slightly greater differentiation (mean Gg^ = 
0.077). 

Nine populations of the Siberian stone pine were stud- 
ied by Krutovskii £ind others (1989) to examine interpopu- 
lation genetic differentiation. Their results indicated only 
slight differences among populations. Using Nei's meas- 
ures of diversity, they determined that on average about 
98 percent of allozyme variation was attributed to intra- 
population variability and only 2 percent to interpopula- 
tion variability (mean Gg^ = 0.022). However, they did 
find a significant correlation between geographical dis- 
tance and genetic differences among populations, but only 
for populations separated by 400 km or more. They also 
found a higher level of differentiation among populations 
within the same region than between regions, which may 
reflect adaptation to local environmental conditions. The 
high level of intrapopulational variation in Siberian stone 
pine is very similar to what has been found for most wind- 
dispersed conifer species (Krutovskii and others 1989). 

EVOLUTIONARY IMPLICATIONS 

Tomback and Linhart (1990) speculated that bird- 
dispersed pines might be more prone to genetic drift than 
are wind-dispersed species. The ranges of these pines, 
particularly subalpine species, tend to be highly frag- 
mented, and small, isolated populations are not uncom- 
mon. These smaller populations may show founder ef- 
fects, particularly loss of rare alleles. In addition, because 
of the occurrence of tree clusters and the potential for 
cross-pollination among cluster members, there may be 
high rates of inbreeding in these populations. When effec- 
tive population size is smaller than 100, genetic drift can 
be far more potent in causing gene frequency changes 
than natural selection (Futuyma 1986 and references 
therein). It is possible that the evolution of cone, seed, 
and tree morphology in bird-dispersed pines was in part 
the consequence of new traits that arose from a combina- 
tion of drift and inbreeding in small, isolated populations. 
Nutcrackers, or ancestral corvids, selected for preferred 
traits and spread these traits to other populations or 
founded new populations with these traits. 

If bird-dispersed pines have this population structure, 
Tomback and Linhart (1990) make two predictions: "(1) 
there is greater genetic variation among populations of 
bird-dispersed pines than among wind-dispersed pines, 
and (2) there are small, isolated populations of bird- 
dispersed pines with highly differentiated gene pools com- 
pared to one another and, particularly, compared to large 
populations that may regularly experience some gene flow 
from other large populations." 

Although the earliest populations of bird-dispersed pines 
may have had the characteristics described by Tomback 



and Linhart (1990), the findings of Krutovskii and others 
(1989), Schuster and Mitton (1992), and Hamrick (unpub- 
lished data cited in Betancourt and others 1991) do not 
support the first prediction, and it may well be that the 
great distances over which nutcrackers disseminate seeds 
provide effective gene flow. Betancourt and others (1991) 
describe a disjunct population of Colorado pinyon pine in 
northern Colorado that was the result of a founding event 
probably about 400 years ago. Compared to other pinyon 
pine populations, the average percentage of polymorphic 
loci and nimiber of alleles per locus were reduced, although 
levels of heterozygosity were not affected. These, how- 
ever, are typical founder effects (Betancourt and others 
1991; Futuyma 1986). Within the last century, the found- 
ing population has given rise to four descendant popula- 
tions, three with fewer than 50 trees each (Premoli and 
others 1993). The two smallest, most distant descendant 
populations had fewer alleles per polymorphic locus and 
reductions in individual heterozygosity compared to the 
larger populations. The mean Gg^ value of 0.060 indicated 
little differentiation among populations. Because tree 
clusters do not occur often in Colorado pinyon pine, these 
new populations may not adequately test prediction 2. 
However, they do demonstrate that seed dispersal by 
birds results in small, isolated foimding populations. 
Much more work is required to thoroughly characterize 
the genetic population structure of bird-dispersed pines 
and to address questions concerning the biology of tree 
clusters. 

ACKNOWLEDGMENTS 

We thank Konstantin Krutovskii and Dmitri Politov for 
providing translations of their papers £ind for interesting 
discussions during the workshop. In addition, we are 
grateful to Jeff Karron, Andrea Premoli, and Kathleen 
Shea for helpful comments on an earlier draft of the pa- 
per. Support for DFT for attending the SSPE Workshop 
was provided by the College of Liberal Arts and Sciences, 
University of Colorado at Denver. 

REFERENCES 

Betancourt, J. L.; Schuster, W. S.; Mitton, J. B.; 
Anderson, R. S. 1991. Fossil and genetic history of 
a pinyon pine (Pinus edulis) isolate. Ecology. 72(5): 
1685-1697. 

Bibikov, D. I. 1948. On the ecology of the nutcracker. 
Trudy Pechorskogo-Ilychskogo Gosudarstvennogo 
Zapovednika. 4(2): 89-112. [In Russian, translated 
by L. Kelso]. 

Bullock, S. J. 1981. Aggregations of Prunus ilicifolia 
(Rosaceae) during dispersal and its effect on survival 
and growth. Madrono. 28: 94-95. 

Carsey, K. S.; Tomback, D. F. 1992. Growth form distribu- 
tion and genetic relationships in tree clusters of Pinus 
flexilis, a bird-dispersed pine. Manuscript. 

Critchfield, W. F.; Little, E. L., Jr. 1966. Geographic 
distribution of the pines of the world. Misc. Publ. 991. 
Washington, DC: U.S. Department of Agriculture, For- 
est Service. 97 p. 



48 



Epperson, B. K 1990. Spatial patterns of genetic varia- 
tion within plant populations. In: Brown, A. H. D.; 
Clegg, M. T.; Kahler, A. L.; Weir, B. S., eds. Plant popu- 
lation genetics, breeding, and genetic resources. 
Sunderland, MA: Sinauer Associates: 229-253. 

Feldman, R. 1991. Growth form and reproductive output 
in limber pine: the cost of mutualism. Denver, CO: Uni- 
versity of Colorado at Denver, Department of Biology. 
46 p. Thesis. 

Feldman, R.; Tomback, D. F. 1991. Growth form and re- 
productive output in limber pine: the cost of mutualism. 
Manuscript. 

Fumier, G. R.; Knowles, P.; Clyde, M. A.; Dancik, B. P. 
1987. Effects of avian seed dispersal on the genetic 
population structure of whitebark pine populations. 
Evolution. 4(3): 607-612. 

Futuyma, D. J. 1986. Evolutionary biology. 2d ed. 
Sunderland, MA: Sinauer Associates. 600 p. 

Hamrick, J. L.; Godt, M. J. W. 1990. Allozyme diversity in 
plant species. In: Brown, A. H. D.; Clegg, M. T.; Kahler, 
A. L.; Weir, B. S., eds. Plant population genetics, breed- 
ing, and genetic resoiirces. Simderland, MA: Sinauer 
Associates: 43-63. 

Hayashida, M. 1989. Seed dispersal by red squirrels and 
subsequent establishment of Korean pine. Forest Ecol- 
ogy and Management. 28: 115-129. 

Holtmeier, F. -K. 1986. Uber Bamninseln (Kollektive) 
an der klimatischen Waldgrenze — imter besonderer 
Beriicksichtigung von Beobachtungen inverschiedenen 
Hochgebirgen Nord-amerikas. Wetter und Leben. 38: 
121-139. 

Hutchins, H. E.; Lanner, R. M. 1982. The central role of 
Clark's nutcracker in the dispersal and establishment 
of whitebark pine. Oecologia. 55: 192-201. 

Keeley, J. E. 1988. Population variation in root grafting 
and a hypothesis. Oikos. 52(3): 364-366. 

Kephart, S. R. 1990. Starch gel electrophoresis of plant 
isozymes: a comparative analysis of techniques. Ameri- 
can Journal of Botany. 77(5): 693-712. 

Knowles, P. 1984. Genetic v£iriability among and within 
closely spaced populations of lodgepole pine. Canadian 
Journal of Genetics and Cytology. 26: 177-184. 

Krutovskii, K. V.; Politov, D. V.; Altukhov, Y. P.; Milyutin, 
L. I.; Kuznetsova, G. V.; Iroshnikov, A. I.; Vorob'ev, V. N.; 
Vorob'eva, N. A. 1989. Genetic variability in the Sibe- 
rian stone pine, Pinus sibirica Du Tour. IV. Genetic 
diversity and the extent of genetic differentiation be- 
tween populations. Genetika. 25(11): 2009-2032. [Trans- 
lation published by Pleniun Publishing Corp. in 1991]. 

Laimer, R. M. 1980. Avian seed dispersal as a factor in 
the ecology and evolution of limber and whitebark 
pines. In: Dancik, B. P.; Higginbotham, K. 0., eds. Sixth 
North Americsin forest biology workshop: proceedings. 
Edmonton, AB: University of Alberta: 15-48. 

Lepper, M. G. 1974. Pinus flexilis James and its environ- 
mental relationships. Davis, CA: University of California 
at Davis, Graduate Group in Ecology. 197 p. Dissertation. 

Ligon, J. D. 1978. Reproductive interdependence of 
pinon jays and pinon pines. Ecological Monographs. 
48: 111-126. 



Linhart, Y. B. 1989. Interactions between genetic and eco- 
logical patchiness in forest trees and their dependent 
species. In: Bock, J. E.; Linhart, Y. B., eds. Evolutionary 
ecology of plants. Boulder, CO: Westview Press: 1-31. 

Linhart, Y. B.; Mitton, J. B.; Sturgeon, K. B.; Davis, M. L. 
1981. Genetic variation in space and time in a popula- 
tion of ponderosa pine. Heredity. 46: 407-426. 

Linhart, Y. B.; Tomback, D. F. 1985. Seed dispersal by 
nutcrackers causes multi-trunk growth form in pines. 
Oecologia. 67: 107-110. 

Mattes, H. 1982. Die Lebengemeinschaft von Tannenhaher 
imd Arve. Berichte Eidenossische Anstalt fur das 
forstliche Versuchswesen, Birmensdorf. Nr. 241. 74 p. 

McCaughey, W. W.; Schmidt, W. C; Shearer, R. C. 1986. 
Seed-dispersal characteristics of conifers in the Inland 
Mountain West. In: Shearer, R. C, comp. Proceedings — 
conifer tree seed in the Inland Mountain West sympo- 
sium; 1985 August 5-6; Missoula, MT. Gen. Tech. Rep. 
INT-203. Ogden, UT: U.S. Department of Agriculture, 
Forest Service, Intermoimtain Research Station: 50-62. 

McNaughton, G. M. 1984. Comparative water relations 
of Pinus flexilis at high and low elevations in the cen- 
tral Rocky Mountains. Laramie, WY: University of 
Wyoming, Department of Botany. 59 p. Thesis. 

Mezhenny, A. A. 1961. Food competitors, enemies and dis- 
eases. In: Egorov, O. V., ed. Ecology and economics of 
the Yakut squirrel. Moscow: Akademiya Nauk: 124-129. 

Politov, D. v.; Krutovskii, K. V. 1990. Genetic variation of 
Siberian stone pine Pinus sibirica Du Tom*. V. Analysis 
of mating system. Genetika. 26(11): 2002-2011. [Trans- 
lation published by Plenimi Publishing Corp. in 1991]. 

Premoh, A. C; Chischilly, S.; Mitton, J. B. 1993. Levels 
of genetic variation captured by four descendant 
populations of pinyon pine (Pinus edulis Engelm.). 
Manuscript. 

Quellar, D. C; Goodnight, K. F. 1989. Estimating relate d- 
ness using genetic markers. Evolution. 43: 258-275. 

Schuster, W. S. F.; Alles, D. L.; Mitton, J. B. 1989. Gene 
flow in limber pine: evidence from pollination phenology 
and genetic differentiation along an elevational transect. 
American Journal of Botany. 76(9): 1395-1403. 

Schuster, W. S. F.; Mitton, J. B. 1991. Relatedness within 
clusters of a bird-dispersed pine and the potential for 
kin interactions. Heredity. 67: 41-48. 

Schuster, W. S. F.; Mitton, J. B. 1992. Gene flow in 
limber pine. II. Data for five populations. Manuscript 
in preparation. 

Swanberg, P. O. 1956. Territory in the thick-billed nut- 
cracker A^uci/ra^a caryocatactes. Ibis. 98: 412-419. 

Tomback, D. F. 1978. Foraging strategies of Clark's nut- 
cracker. The Living Bird. 16: 123-161. 

Tomback, D. F. 1982. Dispersal of whitebark pine seeds 
by Clark's nutcracker: a mutualism hypothesis. Journal 
of Animal Ecology. 51: 451-467. 

Tomback, D. F. 1988. Nutcracker-pine mutualisms: multi- 
tnmk trees and seed size. In: Ouellet, H., ed. Acta XIX 
Congressus Internationalis Ornithologici, Vol. 1; 1986 
June 22-29. Ottawa, ON: University of Ottawa Press: 
518-527. 



49 



Tomback, D. F. 1988. [Unpublished observations in the 
Upper Engadine Valley of Switzerland, October 1988]. 

Tomback, D. F.; Knowles, J. W. 1989. Post-fire whitebark 
pine seed dispersal by Clark's nutcracker in Yellowstone 
National Park. Unpublished data and observations. 

Tomback, D. F.; Linhart, Y. B. 1990. The evolution of bird- 
dispersed pines. Evolutionary Ecology. 4: 185-219. 

Tomback, D. F.; Hoffmann, L. A.; Sund, S. K. 1990. Coevo- 
lution of whitebark pine and nutcrackers: implications 
for forest regeneration. In: Schmidt, W. C; McDonald, 
K J., comps. Proceedings — sjnnposium on whitebark 
pine ecosystems: ecology and management of a high- 
mountain resource; 1989 March 29-31; Bozeman, MT. 
Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Department 
of Agriculture, Forest Service, Intermountain Research 
Station: 118-129. 

Tomback, D. F.; Holtmeier, F. -K.; Mattes, H.; Carsey, 
K. S.; Powell, M. L. 1993. Tree clusters and growth form 



distribution in Pinus cembra, a bird-dispersed pine. 

Arctic and Alpine Research. [In press]. 
Turcek, F. J. 1966. Uber das Wiederauffinden von im 

Boden versteckten Samen durch Tannen-und 

Eichelhaher. Waldhygiene. 6: 215-217. 
Vander Wall, S. B. 1992. The role of animals in dispersing 

a "wind-dispersed" pine. Ecology. 73: 614-621. 
Vander Wall, S. B.; Balda, R. P. 1977. Coadaptations 

of Clark's nutcracker and the pinon pine for efficient 

seed harvest and dispersal. Ecological Monographs. 

47: 89-111. 

Weaver, T.; Jacobs, J. 1990. Occvirence of multiple stems in 
whitebark pine. In: Schmidt, W. C; McDonald, K J., 
comps. Proceedings — symposium whitebark pine ecosys- 
tems: ecology and management of a high-mountain re- 
source; 1989 March 29-31; Bozeman, MT. Gen. Tech. 
Rep. INT-270. Ogden, UT: U.S. Department of Agricul- 
ture, Forest Service, Intermountain Research Station: 
156-159. 



50 



Basic Ecology 




International Workshop 
St. Moritz 1 992 



EXPERIENCES WITH REPRODUCTION 
OF CEMBRA PINE 



Ernst Frehner 
Walter Schonenberger 



Abstract — ^As regards reproduction, cembra pine {Pinus cembra 
L.) is a special case. If the greater part of the seeds are stolen by 
nutcrackers, they should be harvested before they are ripe and 
then matured. With correct treatment it is possible to store 
seeds for up to 7 years. After soaking in water, germination be- 
gan almost at once. The separate steps along the long process 
from harvesting via seed extraction, maturation, storage, testing, 
sowing, and transplantation to planting out are described in 
detail. 



in the Alps, and that the existing stands would diminish 
considerably within a few generations. 

Depending on the number of cones produced, the nut- 
cracker stores seeds as winter provisions in nimierous 
caches, usually numbering over 10,000, each containing 
from three to five seeds, but only consimies about 80 per- 
cent of its reserve. The remaining seeds contribute 
greatly to the regeneration of cembra pine stands. The 
nutcracker may transport seeds over a distance of 15 km 
or more and over a difference in height of up to 600 m. 



This study reports our experiences over many years of 
seed £ind afforestation trials with cembra pine {Pinus 
cembra L.), from seed harvesting to planting. 

Cembra pine is a special case as regards reproduction 
(Frehner and Fuerst 1992). The extreme conditions in its 
natural environment at the upper timberline strongly in- 
fluence flowering, fertilization, and seed maturation, so 
that the quantity and quality of seed varies from year to 
year (table 1). The results of repeated studies on cone 
quality, maturation, and storage vary correspondingly. 

The special features of reproduction in cembra pine be- 
gin with seed collection. Except in full mast years, most 
of the seeds have been eaten or collected by the nut- 
cracker by the time of their natural maturation. Like the 
seeds of many other tree and shrub species, those of 
cembra pine require several months of dormancy. Matu- 
ration of the seed and time of extraction is therefore very 
important for the development of the embryo. To better 
exploit fiill mast years, the durability of the seeds should 
be extended through improved storage methods. It is also 
important that stored seeds should be induced to germi- 
nate and produce successful saplings at short notice. Also 
in terms of sowing, transplantation, and potting, cembra 
pine requires special conditions. 

SYMBIOSIS WITH THE 
NUTCRACKER 

The importance of the nutcracker for the natural distri- 
bution of cembra pine is undisputed. The detailed study 
by Mattes (1982) on this relationship concludes that with- 
out the nutcracker there would be no cembra pine forests 



CONE COLLECTION 

In years with normal, moderate fructification, the nut- 
cracker begins to pick the seeds at medium altitudes as 
soon as ripening has begun, about mid-August, and con- 
tinues to do so until about the end of September. This 
means that collecting cones is a race with the bird. If 
cones are to be collected in years with partial mast, this 
must be done at an early stage, while they are still un- 
ripe, before the birds begin their robbery, and then stored 
under optimvmi conditions for maturation. 

In long-term observations, 19 percent of the years pro- 
duced full mast, 55 percent partial mast, and 26 percent 
failed to produce (Mattes 1982). As a rule it is only in fiill 
mast years that enough seeds ripen on the tree to meike 
harvesting worthwhile; in this case the nutcracker does 
not have time to devour or carry away all the remaining 
seeds. 

Neither color nor size of cones has any influence on seed 
quality or planting success. The costs of harvesting and 
seed extraction are considerably greater for small cones. 

With early harvesting it is very important that the 
seeds should be allowed to ripen imder optimum condi- 
tions of moisture and warmth, as the embryos are in a 
very early stage of development (only up to 3 mm long). 

Studies on protected cones attached to mother trees at 
2,000 m have shown that the length of the embryo hardly 
changes between mid-September and the end of October. 
In contrast, embryos in cones harvested early (mid- 
September) have elongated by 3-4 mm. This shows that 



Table 1 — Characteristics of cembra pine seed crops 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Ernst Frehner is former Head of the Research Nursery, Walter 
Schonenberger is Head of the Research Section Silviculture, Swiss 
Federal Institute for Forest, Snow and Landscape Research, CH-8903 
Birmensdorf, Switzerland. 



Harvesting time per tree 

Yield per tree 

Number of cones per tree 

Weight per cone 

Seed yield (fresh weight) 

Number of living seeds per cone 

Thousand-seed weight 



about 1 hour 
about 15 kg cones 
about 150-200 
60-100 g 
20-30 percent 
50-70 
300-400 g 



52 



Table 2 — Cone collection, maturation, and seed extraction 

September Cone harvesting (embryo length up to 3 mm) 

September Stratification of cones for maturation at 1 0-20 °C 

for 8-1 0 weeks 

November Seed extraction (embryo length 3-5 mm) 



it is not worthwhile enclosing whole trees in nets or treat- 
ing the cones with chemicals. Early harvesting and favor- 
able conditions of stratification are preferable to late 
harvesting. 

CONE MATURATION 

Stratification (storing cones in layers in a moist, airy 
substrate) is the simplest and most certain technique of 
accelerating ripening and reducing inhibition of germi- 
nation so that the seeds are ready to germinate quickly. 
Stratification simulates the natural conditions of moisture 
and warmth prevailing in the forest, or may even be 
better. 

The fi-eshly harvested cones are stored in layers at a 
proportion of one part cones to five parts substrate. Suit- 
able substrates are washed sand or a mixture of sand and 
peat. The substrate must be sieved so that its particles 
are smaller than the seeds. The stratified cones should 
then be stored in a cupboardlike container with good 
drainage that protects the cones from attack by mice or 
birds. 

The development of the embryos depends greatly on the 
temperature of the substrate and the water content of the 
cones. Studies have shown that the embryos develop only 
slowly at low temperatures, while high temperatures also 
have a negative influence. The best results were obtained 
with substrate temperatures alternating between 10 and 
20 °C. It is advisable to keep the stratified cones in a pro- 
tected place (for example, an unheated plastic tiinnel or 
glasshouse), where the temperature may rise toward 
25 °C during the day and sink to about 5 °C at night. 
Naturally, the temperatiire within the substrate does not 
vary so greatly. 

Moisture is very important. The substrate should be 
moist and airy, and the cones should be just moist enough 
to prevent the seed coat from shrinking away from the 
endosperm. Maturation to the point of seed extraction 
proceeds much better under moist conditions than with 
dry storage. With ripening under dry conditions the ger- 
mination rate is 50 percent lower than that after moist 
storage, while dormancy is 10 times longer. 

SEED EXTRACTION 

After 8 to 10 weeks' stratification, the seeds can be ex- 
tracted. The cones, which are by this time beginning to 
disintegrate, are sieved out of the substrate, left to dry for 
a while, and then rubbed through finer sieves to separate 
the seeds from scales and cone axes. The finer scales and 
substrate particles can then be removed through further 



sieving and washing in a water bath, either immediately 
or after another ripening process. In a water bath the im- 
fertilized seeds (empty seed cases) float to the top and can 
be decanted. To summarize, the course of cone matura- 
tion is shown in table 2. 

SEED MATURATION 

The treatment of the seeds after extraction depends on 
their intended use and how long they are to be stored. 

Under natural conditions it takes 2 to 3 years for the 
seeds to germinate after the cones have fallen. This may 
be due to many factors, for example, hard and imperme- 
able seed cases, incomplete development of the embryo, or 
inhibiting substances in the endosperm, in the seed case, 
or elsewhere in the seed. In cembran pine, the main rea- 
son is the low stage of development of the embryos: at the 
time of harvesting the embryos are not more than 3 mm 
long. 

With artificial maturation using variable temperatures, 
the period between harvesting and germination can be re- 
duced to one winter. Here, the aim is to induce the em- 
bryos to grow fi"om a length of 3 mm or less to the germi- 
nation length of 7-9 mm (table 3). To promote embryo 
growth even after seed extraction, the seeds are alter- 
nately exposed to warmth and cold in a monthly rotation. 
Exposure to warmth for periods of more than 4 to 6 weeks 
does not positively influence embryo development. Great 
attention should be paid to water content throughout the 
whole period of seed treatment, as during cone storage. 
The better the seed maturation, the better the embryo de- 
velopment and, consequently, the better the seeds are 
suited for drying and storage. 

BASIC PRINCIPLES OF STORAGE 

After drying, the seeds are placed in one of the con- 
tainer types mentioned and stored imder cool conditions. 
In the early stages they should be regularly inspected. 

Water content: 

• the lower the water content, the lower the storage 
temperature; 

• the lower the storage temperature, the longer the 
seeds can be stored; 

• the longer the seeds have been stored, the longer 
they must be stratified before sowing. 



Table 3 — Results of the maturation of cembran pine seeds 
(harvested September 22) after warm-cold-warm 
treatment 



Features 


September 23 


February 28 


Seed length (peeled) 


9.5 mm 


10.4 mm 


Channel length 


4.5 mm 


8.5 mm 


Embryo length 


1.0 mm 


8.4 mm 


Seeds with visible embryos 


11 percent 


72 percent 


Seeds ready to germinate 


96 percent 


76 percent 



53 



Storage containers: 

• at high water content (>20 percent), in perforated 
plastic bags; 

• at low water content (<15 percent), in airtight con- 
tainers accommodating 2-5 kg. 

Storage rooms: climatized: 

• for short-term storage with high water content, 
1-3 °C; 

• for fairly long storage with low water content, 
-5 to -10 °C. 

Freeze drying has proved unsuitable for cembran pine 
seeds, because the water content is too high. 

Fungal attack: for prevention, the seeds should be 
treated with a fungicide and regularly examined as to 
smell and condensed moisture indicating a too-high water 
content in the containers; if this occurs, the seeds should 
be dried immediately. 

STORAGE OF MATURE SEEDS 

Without treatment for reduction of water content the 
seeds cannot be stored longer than one winter. With re- 
duction of water content to about 25 percent, the seeds 
can be stored at 1-3 °C for up to 3 years without great loss 
in the germination rate. 

Later trials showed that gently handled seeds could be 
dried to a water content of 15 percent, and that under 
storage at -5 to -10 °C the germination capacity did not 
decrease for up to 7 years. A water content of more than 
20 percent rendered storage below 0 °C iinfeasible. 

THREE SEED TREATMENTS 

After seed extraction three treatments for three differ- 
ent storage periods can be defined as: 

• Treatment 1: for immediate sowing without storage 
(table 4). 

• Treatment 2: for storage over 2 to 3 years (table 5). 

• Treatment 3: for storage over 3 to 7 years (table 6). 

GERMINATION TRIALS 

Germination trials begin in December. The seeds are 
sown in a sterile peat/sand in sterilized containers, being 

Table 4 — Treatment 1 : for immediate sowing without storage 

November Seed stratification outdoors for 5-6 months; 

substrate three parts washed sand to one part 
seeds 

April Sowing (usually 30-50 percent of seeds lie-over) 

Advantages: No need for climatization 

Disadvantages: No storage possible, annual seed harvesting 

necessary, high proportion of lie-over, high risk 
of robbery by birds and mice 



Table 5 — Treatment 2: for storage over 2-3 years with high water 

content under refrigeration in plastic bags (for example, in 
years with partial or full mast) 



November Seed stratification for 4-6 weeks at 1 6-1 8 °C in 

washed sand 

December/January Sieving and cleaning of seeds; drying of seeds 
for 1 -2 weeks at 1 6-1 8 °C to about 
25 percent water content: embryo length 
increases to 5-7 mm 

January Storage of seeds in perforated plastic bags at 

1 -3 °C over 2-3 years 

October Seeds are soaked for 1 -2 days; stratification 

outdoors for 6 months in washed sand 

April/May Sowing (no lie-over to be expected) 



Advantages: Allows storage for up to 3 years without lie-over 

Disadvantages: Storage climatized at 1-3 °C necessary, high 
water content, risk of rotting and fungal 
infection, long stratification period 



normally spread and pressed down, then covered with 
about 1 cm sand or substrate. The trials are best con- 
ducted in rooms without direct sunlight; the ideal tem- 
perature is 16 to 18 °C. Alternating temperatures have 
a positive effect. The seeds should be watered by hand 
according to need. Storage in greenhouses with automatic 
spraying has not proved suitable. The germination tests 
take 6-8 weeks. 



Table 6 — Treatment 3: for storage in airtight containers for 3-7 years 
with low water content and freezing temperatures (for 
example, for seeds harvested In good partial or full mast 
years) 



December 


Seed stratification; first warm treatment at 




16-18 °C for 4-5 weeks 


January 


Cold treatment at 1 -3 °C for 4-5 weeks 


February 


Second warm treatment at 1 6-1 8 °C for 




4-5 weeks until germination begins; embryo 




length increases to 7-9 mm 


February 


Sieving and cleaning of seeds; drying at 




1 6-1 8 °C for 1 -2 weeks tol 5 percent water 




content 


March 


Storage in airtight plastic bags or glass 




containers at -5 to -1 0 °C for 3-7 years 


April 


Defrosting for 1-2 days, soaking in lukewarm 




water for 1-2 days, sowing 


Advantages: 


Seeds can be stored for up to 7 years, no 




lie-over, short-term stratification (6-8 weeks). 




good mast years and seed quality can be 




fully exploited 


Disadvantages: 


Long-term deep-freezing required 



54 



SOWING 

The seeds are sown between the end of April and mid- 
May. The number of seeds to be sown depends on the re- 
sults of the germination trials. Because germination is 
slow, the seeds should be sown in boxes resistant to mice 
and birds. The best substrates are soils without lime- 
stone mixed, according to conditions and pH, with peat 
(which loosens the soil, maintains a pH between 5 and 6, 
and regulates the moisture level) or raw humus contain- 
ing mycorrhizal fungi from cembran pine stands. The 
seed and transplantation beds should continuously be 
occupied by cembran pines except for short periods with 
green fertilizing. 

Cembran pine grows best in thickly sown beds, with 
1,200 to 1,800 seedlings per square meter, grown for 2 
to 3 years in a seedbed. 

If the seeds begin to germinate during stratification, 
they should be sown immediately. If this is unfeasible, 
they can be frozen at -3 to -4 °C and sown as required. 
Seeds that have already germinated and whose radicle is 
longer than 1 cm should not be planted in a seedbed, but 
thinned out; otherwise the resulting plants are liable to 
be worthless, with tortuous root collars, short stems, 
small crowns, and few fine roots. Size, quality, and yield 
are definitely better in thinned seedlings. Shortening the 
roots causes deformations. 

We did not run any trials with sowing in autumn, as 
the risks of attack by mice and birds and also snow pres- 
sure on the protection lattice are then very high. 

TRANSPLANTING 

The specially designed beds are continuously supple- 
mented with acid soil (for example, peat) and if necessary 
with chalk-free acid fertilizer. To hinder invasion by 
weeds and frost heaving and to restrict evaporation, and 



also to maintain a good soil structiire, the beds are 
mulched. 

It is recommended that 3-year-old saplings be trans- 
planted in spring or summer (from mid-July). This can be 
done by machine, with a spacing of 10 by 18 cm. Saplings 
with poor root development should be discarded, as it is 
not worth planting them out. The saplings stay in the 
transplanting bed for 3 years. 

POTTING 

High-altitude afforestations are exposed to high risks 
from extreme weather conditions, snow, and pests. Expe- 
rience has shown that seedlings in peat-fiber pots succeed 
better than those with naked roots, especially on diffic\ilt 
sites, so potting is recommended. The production of pot- 
ted plants has been described by Schonenberger and 
others (1990). It is important to keep the pots well 
watered. The saplings should only be planted out when 
the root tips begin to penetrate through the peat-fiber 
pots. 

REFERENCES 

Frehner, Ernst; Fiirst, Ernst. 1992. Vom Samen bis zur 
Pflanze. Ein Erfahrungsbericht aus dem Forstgarten. 
Ber. Eidgenoss. Forsch. anst. Wald Schnee Landsch. 
333. 47 p. 

Mattes, Hermann. 1982. Die Lebensgemeinschaft von 
Tannenhaher und Arve. Eidg. Anst. forstl. 
Versuchswes., Ber. 241, 74 S. 

Schonenberger, Walter; Frey, Werner; Leuenberger, 
Franz. 1990. Oekologie und Technik der Aufforstung 
im Gebirge - Anregungen fiir die Praxis. Eidg. Anst. 
forstl. Versuchswes., Ber. 325, 58 S. 

Translation: Margaret Sieber 



55 



DISTRIBUTION AND ECOLOGY OF 
SIBERIAN STONE PINE IN THE URALS 



P. L. Gorchakovsky 



Abstract — The northern and southern boundaries of the distri- 
bution area of Siberian stone pine (Pinus sibirica) in the Urals 
and in the adjacent plains are examined; the main sociological 
factors limiting boundary positions are described. There is a 
natural tendency of this species to spread northward and south- 
ward, from places recently invaded, but this process is usually 
stopped or turned back by human influence. Numerous occur- 
rences of Siberian stone pine beyond its present continuous range 
indicate that the former distribution area of this species was 
larger in the past than at present. 



Siberian stone pine (Pinus sibirica), or briefly Siberian 
pine, is considered to be one of the most valuable species 
of coniferous trees growing in Russia. Research on the 
limits of natural distribution of Siberian pine, on their 
causes and dynamics, is of indisputable theoretical and 
practical interest. Research is especially necessary for 
scientific substantiation of rational use of all forests, par- 
ticularly those with Siberian pine, for the definition of 
measures to protect it fi"om being exterminated, and for 
solving the problems concerning possible cultivation of 
this species in adjacent regions. 

The general distribution area of Siberian pine includes: 
(1) northeastern regions of the European part of Russia 
(the farthest west locality is in the upper reaches of the 
Vychegda River); (2) the Ural Mountain range; (3) west- 
em Siberia; (4) middle and eastern Siberia (up to the 
Aldan River); and (5) mountain ranges of Altai, the Trans- 
baikal area, and northern Mongolia. 

The main limiting factors of the distribution of this 
species are: heat deficiency in the north; permafi*ost, peat- 
Ismd, and severe winter in the northeast; water deficiency 
and low air hvunidity in the south. 

The western part of the area is located in the Ural 
Mountains and in the northeastern regions of the Euro- 
pean part of Russia adjacent to them. Siberian pine occu- 
pies rather large areas within the territory of Sverdlovsk 
and Perm provinces, Komi Republic, and within Knanty- 
Mansy national district of Tyimien province. 

THE NORTHERN LIMIT 

The northern border of the distribution area (fig. 1) 
passes along the Izhma-Pechora watershed, advances 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

P. L. Gorchakovsky is Professor, Doctor of Biology, Chief Scientist at 
the Institute of Plant and Animal Ecology, Russian Academy of Sciences, 
Ural Branch, 620219, Ekaterinburg, Russia. 



considerably northward (up to 65°30' N. latitude) along 
the valley of the Pechora River and even farther north- 
ward (67° N. latitude — the northern polar circle). Then 
it switches back south in the form of a tongue, to the foot- 
hills of the elevated eastern part of the Russian plain. 
From Telpos-Iz Mountain, 64° N. latitude, it leaps north- 
ward along the eastern slope of the Poleir and Prepolar 
Urals reaching 66° 10' N. latitude. It then recedes south- 
ward in the space between the rivers, but returns north- 
ward again in a tongue-shaped projection along the vsdley 
of the Ob River. 

Besides that, a few isolated occurrences of Siberian pine 
are known outside its main distribution area. 

The northern limit of Siberian pine does not reach the 
extreme boundaries of the distribution of larch (Larix 
sibirica) and spruce (Picea obovata). The well-known 
Russian botanist B. N. Gorodkov (1929) supposed that the 
advancement of Siberisin pine to the north is restricted be- 
cause its seeds do not get ripe under the severe conditions 
of the short growing season. 

However, the latest observations demonstrate that 
even at the northern limit of the distribution area seeds 
of Siberian pine usually get ripe. The main reason likely 
is that Siberian pine is not able to develop its root system 
imder low soil temperature (Tyrtikov 1954). In cold soils, 
the roots of Siberian pine ramificate comparatively weakly 
and grow more slowly than the roots of spruce and larch. 
Consequently, insufficient soil heating during the growing 
season impedes normal development of root systems, and 
this appears to be the main factor limiting the distribu- 
tion of Siberian pine in the north. Under the severe cli- 
matic conditions of northern Siberia, pine does not grow 
on cold, peatland soils. 

Close to the northern boundaries, in the upper reaches 
of the Pechora River, Siberian pine selects the warmest 
well-heated locations, inhabiting even the riverside lime- 
stone outcrops and cliffs. Contrary to that, in the regions 
farther south, for example in the middle and southern 
Urals, Siberian pine never grows in such habitats because 
the riverside limestone cliffs are too warm during daylight 
periods. 

In the plains adjacent to the Urals (the Pechora and 
western Siberian lowlands), the far-north occurrences of 
Siberian pine are restricted to the valleys of the large riv- 
ers (the Pechora, the Ob, and some of their tributaries). 
The bovmdary deviates northward in the form of narrow 
strips along the river valleys. In marshy areas between 
the rivers, however, where superficial bedding of perma- 
fi'ost fi*equently occurs, the border line is farther south. 
Under plain conditions, the river valleys characteristically 
have more permeable ground, lower levels of permafi"ost 
bedding, and warmer soils that are more favorable to 



56 



Figure 1 — Northern distribution limit of Siberian stone pine in tine Urals and in the adjacent 
plains. * = isolated occurrences. 



Siberian pine. Moreover, one should bear in mind that 
the valleys of the large rivers that bring masses of 
comparatively warm water from the south are character- 
ized by a warmer local climate favorable to the growth of 
Siberian pine. 

At the Lyapin depression, in the region of Saranpaul, 
Siberian pine- dominated forests are to be found on buried 
palsa peatbogs covered by sandy loam deposits. On hill- 
ocks, better conditions for drainage and heating of the su- 
perficial layer of accumulated deposits favor the formation 
of rather tall Siberian pine forests. On the other hand, 
raised bogs with pools and ridges between the hillocks are 
treeless. The level of permafrost here is rather high; ice- 
lenses are foimd inside the peaty hillocks. 

In the plains, at the extreme Hmit of its distribution 
area, Siberian pine is usually represented by solitary 
specimens dispersed within spruce-larch-birch forests; 
trees usually are healthy, without signs of atrophy; their 
tnmks reach 10 to 14 m in height. 

Within the western and eastern outskirts of the Ural 
Moimtain range, the Siberian pine border displaces north- 
ward along meridionally oriented low foothills and moim- 
tain ridges. There, this species grows on slopes or near 
brooks and small rivulets imder moderate or slightly in- 
creased (flowing) ground moisture conditions. The advance 
northward is promoted by a variety of sites suitable for for- 
est growth as well as by favorable drainage, better soil 
heating, and consequently by lower depth of permafrost. 

In the watershed of the Urals there are huge mountains 
topped by woodless tundrahke "goltsy" summits, located 



within the mountain tundra and high-mountain cold 
desert belts (Gorchakovsky 1975, 1989). They are char- 
acterized by a more severe climate, and consequently the 
northern limit of Siberian pine is considerably displaced 
southward. 

Siberian pine almost absolutely escapes the moimtain 
massif of the Prepolar Urals from the west; its distribu- 
tion is restricted to the Pechora lowland; in rare cases it 
occurs in the moimtain part along deep erosion-tectonic 
valleys of the large rivers. 

On the eastern slope of the Prepolar Urals, however, 
Siberian pine ascends rather high in the mountains, al- 
most to the upper forest limit. In the "subgoltsy" belt 
(analogous to the subalpine belt), in the region of Gorodkov 
Mountain (Sale-Ur-Oika), Siberian pine shows dwarfed 
growth (2.5 to 5 m in height), exhibiting trunks thickened 
at the basal portion of the tree and tapering upward, and 
with wide-branched crowns of low stability. Frequently, 
such dwarfed trees bear abundant cones. At lower eleva- 
tions, within the mountain-taiga belt in the plain, Sibe- 
rian pine reaches 10 m in height. From the eastern side, 
along the foothills, this plant extends to the Polar Urals 
reaching 66° 10' N. latitude in the mountains of the so- 
called Small Urals. 

More intensive advancement of Siberian pine north- 
ward, along the eastern slope of the Ural Mountain range, 
was favored partly by historical causes. This species in- 
vaded from Siberia to new places in the Urals. It arrived 
earher on the eastern slope of the Ural Mountain range, 
favored mainly by the specific natural conditions of the 
eastern slope. The eastern slope of the Prepolar and 



57 



Polar Urals has a more continental climate than the west- 
ern slope. Relative air humidity is lower here; annual 
precipitation including that in the winter is less heavy; 
and snow cover melts earlier in spring. Consequently, the 
growing season is longer, and the top soil layer is better 
heated. 

On its northern boundaries, Siberian pine is distributed 
disjimctively and occurs as dispersed small clumps or soli- 
tary specimens in the taiga. All these extreme localities 
are rather disjimct from each other. This may be ex- 
plained primarily by the fact that only in rare locations 
are the combinations of various environmental factors 
favorable for growth of Siberian pine at the limits of its 
range. On the other hand, the dispersion of this species 
has been reduced here and there due to forest fires, exten- 
sion of bogs, trampling of seedlings and young growth 
by reindeer, and extermination by humans. In locations 
where human activity is reduced to a minimum, Siberian 
pine invades new places north of its present limit. This 
advance is observed especially clearly in the mountain 
regions of the Urals. 



THE SOUTHERN LIMIT 

The southern borderline of continuous distribution of 
Siberian pine (fig. 2) passes through the Russian plain 
near the city of Perm; then it turns southward along the 
western slope of the middle Urals, reaching the most 
southern position at 56° 10' N. latitude; then it switches 
back north following the eastern slope of the middle Urals 
until retiring into western Siberia. 

A few outliers occur beyond this limit. The most inter- 
esting of them are to be found in the mountain dark conif- 
erous taiga of the southern Urals in the region of the town 
of Zlatoust and in the Ilmen Mountains. 

At the southern limit of its range, Siberian pine usually 
grows in the form of single-standing, tall trees. Only on 
rocky outcrops does it exhibit prostrated growth forms 
(krummholz). 

Natiiral regeneration of Siberian pine near the southern 
boimdary of its distribution is generally satisfactory. How- 
ever, there are many factors unfavorable to regeneration, 
such as seed collection difficulties, fires, and trsimpling of 




Figure 2— Southern distribution limit of Siberian stone pine in the Urals 
and in the adjacent plains. * = isolated occurrences. 



58 



seedlings by livestock. In locations where such vmfavor- 
able influences are absent, or they manifest themselves 
to a lesser degree, regeneration of Siberian pine is usually 
good. So, for example, Siberian pine reproduces very suc- 
cessfully on raised sites protected from fire within the 
Bakhmet Bog, which is surroimded by Scots pine forests 
in the basin of the Pyshma River. 

Along the western slope of the Urals, as well as in cer- 
tain sites within the watershed zone where the climate is 
more humid, Siberian pine has invaded new places signifi- 
cantly farther south than in the Preural and Transural re- 
gions. Its southern limit projects southward in the form 
of tongues that comprise the western slope and part of the 
watersheds of the moimtain range. Climate of the eastern 
slope of the middle Urals is drier; the amoiint of precipita- 
tion is less compared to the western slope. Therefore, on 
the eastern slope of the movmtain range and in the foot- 
hills zone adjacent to it (the region of predominance of 
Scots pine forests), Siberian pine occurs considerably far- 
ther north. 

Within the peatland between the rivers of the western 
Siberian plain, the southern limit of Siberian pine turns 
south again. There Siberian pine grows mainly on raised 
sites within bogs, where it is well protected from fire. 

Siberian pine invaded the southern Urals later than 
spruce and fir {Abies sibirica) and has not had enough 
time in the prehistoric period to reach the climatically 
influenced limit of its distribution. Human activities 
hampered the southward advance of Siberian pine into 
new places along the Ural Mountain range. Here and 
there, numbers of this species have declined. Several cen- 
turies ago Siberian pine had been distributed farther 
south in the Urals compared to its contemporary limit. 
Since then it has receded northward due to forest fires 
and himian impact. There is clear evidence of recent de- 
cline in total distribution. In the southern part of the 
area, Siberian pine occurs sporadically in small groves, 
groups of trees, or as single trees. It grows in places least 
endangered by fire, most frequently on moist soils in the 
headwater regions and valleys of brooks and rivulets, on 
the margins of bogs and on raised sites within them, on 
lake shores (somewhere close to water), and on steep out- 
crops of acid rocks. 

Formerly, the region of continuous distribution of Sibe- 
rian pine extended south at least to the isolated localities 
mentioned earlier, but probably this tree species extended 
even farther south. Literature data as well as oral state- 
ments by local residents provide evidence of it. Siberian 
pine disappeared from these places not only due to cutting 
of timber and forest fires, but also because local inhabit- 
ants picked all the cones, often before cones were com- 
pletely ripe. They also dug up seedlings for transplanting 
into personal gardens. 

DISCUSSION AND CONCLUSIONS 

The rather complicated picture of distribution of Sibe- 
rian pine in the Urals depends to a large extent on cli- 
matic difl'erences between the relatively himiid western 
slope of the range and the more continental eastern slope. 



Humidity of the western slope restricted the northward 
advance of Siberian pine, but favored its advance to the 
south. On the contrary, continentality of the eastern 
slope favored the advance of this species northward, but 
restricted its advance southward. 

At both the northern and southern Umits of its distribu- 
tion, Siberian pine does not occur high up in the moun- 
tains. Its northernmost occurrences in high mountains 
are in the eastern part of the Prepolar Urals. 

In the northern Urals, Siberian pine is widespread al- 
most everywhere in the high-mountain belts. On certain 
mountains (Oika-Nyor, Pas-Nyor, Chistop, and others), 
however, it forms the upper forest limit. High-mountain 
Siberian pine-dominated forests grow on steep slopes with 
poorly developed soils. Near the upper forest limit, such 
forests may be foimd mainly in less high mountains far 
removed from large mountain massives. In such isolated 
mountains, the upper forest limit is caused chiefly by 
edaphic factors, while the cHmate of their treeless zone 
and of the subgoltsy belt is less severe. Dwarfed, pros- 
trate, and, in most cases, noncone-producing specimens 
of Siberian pine of the northern Urals occur in high- 
moimtain tundras significantly higher in elevation than 
the normal upper forest limit. For instance, dwarf speci- 
mens occur on the movmtains Isherim, Oika-Chahl, 
Yalping-Nyor, Denezhkin Kamen, Konzhakovsky Kamen, 
Kosvinsky Kamen, and others. The occurrences of kmmm- 
holz growth forms above the normal upper forest limit 
have to be explained by the dispersal of seeds into the 
moimtain tundra zone by the nutcracker and rodents. 

In the middle Urals near the simamits of relatively high 
moimtains, Siberian pine also forms singly standing, 
dwarfed trees (the mountains Kachkanar, Starick-Kamen, 
and others). South of Moimt Starick-Kamen, this species 
is never foimd in high-movmtain belts, but in the forest 
belt its area spreads considerably farther south. 

Judging by its biological and ecological features, Sibe- 
rian pine may not be considered as a species becoming 
extinct. It is perfectly well adapted to the taiga environ- 
ment; it is a good cone producer, and, if human impact 
does not interfere, it regenerates quite satisfactorily. Al- 
most everywhere, regeneration occurs close to old trees. 
However, one should bear in mind that Siberian pine 
grows more slowly than other shade-tolerant coniferous 
trees. Therefore, ground fires that annihilate second- 
growth trees adversely afl'ect natural regeneration of this 
species. 

The border lines of Siberian pine, especially on the 
southern border, have been reduced by careless and some- 
times barbarian treatment of this valuable plant. It is 
necessary to secure proper protective measiires for its 
southern localities in the Urals and to use seeds gathered 
from trees mainly for sowing under the forest canopy and 
in tree nurseries. 

In the coTorse of intensive harvest cuttings in the dark 
coniferous fir-spruce forests that have individual Siberian 
pine trees in the admixture, foresters usually retain Sibe- 
rian pine as seed trees for natural regeneration. How- 
ever, in many cases, this does not work well because in 
the heavy clay and loamy soils of the plains, and in the 



59 



mountains on poorly developed soils, such trees r\in the 
danger of windfall. To avoid windfall, it is necessary to 
retain clumps of Siberian pine with accompanying tree 
species instead of single specimens of Siberian pine. 

Natural conditions of the middle Urals and of the 
moiintain-range portion of the southern Urals (at least 
up to 54° N. latitude) are favorable for regeneration and 
growth of Siberian pine. This valuable tree species may 
be reintroduced to areas south of its present distribution. 
This can be concluded from the successful cultivation of 
Siberian pine south of its present border line in the moun- 
tain taiga regions of the middle and southern Urals. 



REFERENCES 

Gorchakovsky, P. L. 1975. The plant world of the high- 
mountain Urals. Moscow: "Nauka" Publishing. 284 p. 
[In Russian]. 

Gorchakovsky, P. L. 1989. Horizontal and altitudinal 
differentiation of the vegetational cover of the Ural 
Mountains. Pirineos, N 133: 33-54. 

Gorodkow, B. N. 1929. The Polar Urals in the upper 
reaches of the rivers Voykar, Synya and Lyapin. In: 
Materials of the Committee for Expeditional Re- 
searches, V. 7, Leningrad: 1-28. [In Russian]. 

Tyrtikov, A. P. Growth of tree roots at the northern limit 
of forests. Bull, of Moscow Soc. of Naturalists. Section 
of Biology. 59(1): 72-82. [In Russian]. 



60 



ECOPHYSIOLOGICAL INVESTIGATIONS 
ON CEMBRAN PINE AT TIMBERLINE IN 
THE ALPS, AN OVERVIEW 

Rudolf Hasler 



Abstract — During the nineteenth century foresters as well as 
scientists became interested in learning more about cembran 
pine (Pinus cembra) in the Alps, but extensive ecophysiological 
fieldwork did not begin before the 1930's. Great progress was 
made after the estabUshment of permanent field stations at the 
timberUne at Obergurgl (1953), Patscherkofel near Innsbruck 
(1963), and Stillberg near Davos (1959). In a brief overview, 
some selected results of ecophysiological research in cembran 
pine are presented. 



The old and bizarre cembran pines (Pinus cembra L.), 
which form the timbedine over a wide range in the Cen- 
tral Alps, have always fascinated himaans. In earlier 
days, the tree was of some importance. It produced not 
only wood but also nuts that could be used as food. There- 
fore, this tree has been of some interest for a long time, 
and a lot of books and papers about cembran pine in the 
Alps have been published, in forestry as well as in biologi- 
cal science. 

This short overview presents only some selected ex- 
amples of the ecophysiological research, with a special 
focus on gas exchange. A general overview on ecophysio- 
logical problems at the timberline was published by 
TranquilUni (1979). Many other subjects, such as growth, 
genetics, and mycorrhiza, have been investigated in 
Austria, France, Italy, and Switzerland but are not men- 
tioned in this overview. 

BEGINNING OF INVESTIGATIONS 

In the middle of the nineteenth century, foresters be- 
came interested in the cembran pines growing in the Alps. 
Fankhauser (1853) described the distribution of these 
trees together with the climatological conditions they 
need for growing well. He mentioned that it is not 
possible to transfer this tree species to low elevations. 
Kemer (1866) and Simony (1870) pubhshed similar data 
for parts of Austria. At the same time, the diseases and 
the insect enemies of the cembran pine in the Alps at- 
tained some importance. They have been described by 
Keller (1890) and others. 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Rudolf Hasler is Head of the Section "Forest and Climate," Swiss Fed- 
eral Institute for Forest, Snow and Landscape Research, CH-8903 
Birmensdorf, Switzerland. 



ECOLOGICAL AND GEOGRAPfflCAL 
OBSERVATIONS 

More and more scientists became interested in cembran 
pine at the end of the nineteenth century and the early 
twentieth century to learn more about these trees, which 
grow on the most extreme sites in the mountains. Mono- 
graphs concerning cembran pine in Switzerland (RikH 
1909) and in Austria (Nevole 1914) were pubhshed. Many 
other publications were written at this time. The main 
subjects were growth, and again fungal diseases and prob- 
lems with insects (Fankhauser 1903; Keller 1901, 1910; 
and others). 

FIRST ECOPHYSIOLOGICAL FIELD 
EXPERIMENTS 

Ecophysiological investigation in the field began in the 
thirties on Patscherkofel near Innsbruck (Austria). To 
work near the timberline in the Alps was a logistic prob- 
lem. The cembran pines only grow near the timberline, 
at high altitudes. It was extremely hard to work there in 
earher times. 

Cartellieri (1935) was probably the first to measiu*e 
photosynthesis and water relations throughout a whole 
season in cembran pine at timberline. At the same time 
he also took measurements of different dwarf shrubs at 
high elevation. He measured photosynthesis by absorbing 
the carbon dioxide in barite-water, and then he deter- 
mined the COg assimilation by titration in the laboratory. 
Some idea about the difficulties of working at the timber- 
line in those days is given in his papers: missing data in 
the annual course are explained as "walking back to the 
laboratory I lost some of the probes because some of the 
glass flasks were broken," or in another example, "at this 
time I got ill and had to stop the investigation." Never- 
theless his results show more or less the same values as 
we are measuring today with ovu* modem climatized and 
very expensive instrumentation. Also about the same 
time Ulmer (1937) investigated the annual course of fi*ost 
hardiness and measured in parallel the osmotic values 
in the needles. 

PERMANENT FIELD STATIONS AND 
CONTINUOUS CO2 MEASUREMENTS 

Great progress in timberline research in the Alps was 
made in the 1950's. Again in Austria, a permanent 
experimental field station at an altitude of about 1,950 m 
above sea level (a.s.l.) was established in 1953 near 



61 



Obergurgl. At the same time, the newly developed infra- 
red gas analyzer (IRGA) made it possible to measm'e 
continuously carbon dioxide in the air. 

Tranquillini (1955) conducted experiments at the new 
field station at Obergurgl to investigate the relations be- 
tween light, temperature, and photosynthesis in needles 
of young cembran pines growing in the sun or in the shad- 
ows. He determined the light compensation points for 
these needles as 500 and 150 Lux, respectively, and the 
light saturation points as 25 and 18 kLux, respectively. 
The COg assimilation values he reported reached 2.5 mg 
and 2.1 mg COJg dry weight, respectively. Net photo- 
synthesis starts at -5 °C and has an optimum range be- 
tween 10 and 15 °C. The upper temperature limit is at 
35 °C. 

The microclimatic conditions during wintertime and 
needle and soil temperatures as well as gas exchange and 
water relations in young cembran pines growing on natu- 
ral sites at the timberline were published by Tranquillini 
(1957). He showed that in wintertime, even in warmer 
weather conditions, no net photosynthesis was detectable. 
Holzer (1958) explained this lack of photosynthesis by a 
changed cell structure as well as by a changed physiolo- 
gical behavior during wintertime. 

Pisek and Winkler (1958) measured the annual course 
of the capacity of net photosynthesis and of respiration in 
detached shoots of cembran pines growing at different al- 
titudes. During wintertime, they found an always nega- 
tive carbon dioxide balance in the trees at the timberline, 
whereas those growing at lower elevations could produce 
a positive one, at least during periods with temperatures 
above the freezing point. Keller (1970) confirmed these 
findings, using another method. Tranquillini (1958, 
1963d) investigated the annual course of frost resistance. 
He found that young trees are more sensitive to frost than 
mature trees. Young trees become damaged by tempera- 
tures between -15 to -30 °C, while mature trees can sur- 
vive temperatures lower than —45 °C. Keller and Beda- 
Puta (1973) analyzed stem respiration in cembran pine 
during wintertime. 

Pisek and Winkler (1959) studied photosynthesis in re- 
lation to temperature in different light conditions as well 
as at different altitudes. With increasing light intensity, 
the temperature optimum is shifted to higher tempera- 
tures. There are some difficulties in comparing all these 
older results with the measurements of today, because not 
all parameters can be transformed (for example, illumina- 
tion [Lux]). Some other values, such as the needle sur- 
faces, were normally not used as a reference for gas ex- 
change measurements before the 1980's. 

A COg budget for a whole year was estimated by 
TranquiUini (1959a, 1959b, 1963a, 1963b). Total CO^ 
uptake per year (gross photosynthesis) of a 5- to 8-year- 
old tree at 2,000 m a.s.l. was about 7.8 kg. About 38 per- 
cent was lost by respiration from needles, branches, stem, 
and roots. According to the measured net COg assimi- 
lation, a tree should produce 2.2 g dry matter per gram of 
needles. In reality they produced 0.65 g. The author 
mentioned that most of the missing carbon dioxide may 
have been used by the mycorrhiza. 

In the second half of the 1950's, ecological investiga- 
tions with the aim of finding methods for successful 



afforestation at high altitude were started in Switzerland 
(Stillberg, Davos: 2,000-2,230 m a.s.l.) and in Austria 
(Haggen, Sellraintal: 1,730-1,900 m a.s.l.). Cembran pine 
was one of the species used in these field experiments, 
which still are going on. It takes a long time to get results 
from trees growing at high elevation near the timberline. 
Svmimaries containing prelimin£iry results were published 
by Schonenberger and Frey (1988) for Stillberg and by 
Kronfuss (1980, 1983, 1986) for Haggen. 

LABORATORY CLIMATE CONTROL 

The technical possibilities developed fast. On 
Patscherkofel (1,950 m a.s.l.) a climate laboratory was 
established in 1963. In this laboratory a device was in- 
stalled that allows analysis of the gas exchange of yoimg 
trees in climatized windchannels at high altitude. Now it 
became possible to control exactly the climatic parameters 
such as temperature and air and soil humidity as well as 
windspeed, which was a brand-new innovation. Using 
this new instrumentation old measurements could be ve- 
rified and a lot of new possibilities were opened up. In 
particular, water relations could now be investigated un- 
der controlled climatic conditions. 

Tranquillini (1963c) investigated the CO^ assimilation 
in relation to air and soil humidity. In water-saturated 
soil, dry air (25 percent relative hmnidity) reduced photo- 
synthesis in cembran pines by about 33 percent. A little 
less water in the soil (half of field capacity) allowed some 
higher photosjnithetic rates. Resistance to dry himiidity 
increases from spring to autimin (Tranquillini 1965). 

Pisek and others (1967, 1968, 1969) explored the tem- 
perature limits of photosynthesis to chill and frost, and to 
high temperatures. COg assimilation is operative between 
-4.7 and +36 °C, but the trees can survive temperatures 
between below -50 and up to 48 °C. Of great importance 
is not only the time factor, but also the physiological con- 
dition of the plant. Temperature hardiness changes with 
the seasons, £md young trees or seedlings are more sensi- 
tive to extreme temperatures than mature trees. The re- 
lation between photosynthetical capacity and temperature 
was analyzed by Tranquillini and Machl-Ebner (1971). 
Another factor investigated was day length. Bamberg 
and others (1967) and Schwarz (1970a, 1970b, 1971) have 
shown that day length does not influence the photosyn- 
thetic capacity during winter depression, but trees kept 
in warm conditions during winter commenced growth in 
spring much earlier than plants in nature. The variation 
in frost resistance of the trees changed in parallel with 
the temperature. In long-day regimes, frost resistance of 
cembran pine remained at a lower level than those grow- 
ing imder short-day conditions. The researchers supposed 
that an endogenous rhythm also influences resistance to 
extreme temperatures. 

Caldwell (1970), using the climatized wind tunnel of the 
Patscherkofel laboratory, has shown the effect of wind on 
stomatal aperture, photosynthesis, and transpiration. 
He did not detect a great influence of low windspeeds on 
these parameters. At higher windspeeds (up to 8 m/sec) 
photosynthesis and transpiration were slightly reduced. 
Baig and others (1974) and Baig and Tranquillini (1980) 
investigated the cuticular transpiration rate of cembran 



62 



pine and found that cuticular transpiration in this tree 
is very slow but, as in other species, cuticular resistance 
decreases in paredlel to the elevation at which the shoots 
were collected. 

Havranek (1972) found that cold soil temperatures 
reduce photosynthesis and transpiration in cembran pine. 
Havranek and Benecke (1978) investigated the influence 
of soil moistiire on transpiration and photosynthesis. 
Cembran pine begins to close the stomata at a soil water 
potential of about -0.4 bars. Compared to other species, 
cembran pine maintains a high water-use efficiency and 
uses limited soil water economically. Giinthardt and 
Wanner (1982) showed that needles from timberline trees 
have on the average more wax than those grown at a 
lower elevation. 



Another step to get more precise resvilts for the CO^ 
balance of cembran pine was taken by Tranquillini and 
Schiitz (1970). They measured bark respiration at the 
timberline and reported that b£irk respiration also shows 
an annual course. In general, bark respiration of cem- 
bran pine was rather high compared to that of other spe- 
cies. The loss of carbon dioxide by stem respiration 
throughout a year was calculated to be about 18.5 percent 
of gross photosynthesis. Measuring stem respiration con- 
tinuously, Havranek (1981) found that there is a marked 
fluctuation in stem respiration during the growing season, 
but he could not find any relation between activity of stem 
respiration and tree ring growth. 



Net photosynthesis 

40- 

■a 32" 
E 2411 

o 16 
E 81 
E 

-8 




I I I I I B I I I I I 11 I I I I I 11 I I I I I 11 I I I I I 11 I I I I I I I I I I I 11 I I I I I H I I I I I 11 I I I 

5 15 2515 15 2515 15 2515 15 25 15 15 25 1 5 15 251 5 15 2515 15 2515 15 25 I 5 15 
Photosynthetically active radiation 




'l I I I I I 11 I I I I I 11 r r I I I 11 I I 'V r i iTT i i i i ii T i i i i i i i i i i-ir i i i i i n i i i i i ii i i r 
15 15 2515 15 2515 15 251 5 15 25 I 5 15 25 I 5 15 2515 15 25 I 5 15 2515 15 25 I 5 15 

Air temperature 




5 15 2515 15 2515 15 251 5 15 25 I 5 15 25 1 5 15 2515 15 2515 15 251 5 15 25 I 5 15 



Soil temperature 



I M III I I I I I I I I I I I I I I I I I I 11 I I I I I 11 I I I I I I I ( I I I 11 I I I I I 11 I I I I I 11 I I !' 

1 5 15 2515 15 25 1 5 15 2515 15 25 1 5 15 2515 15 251 5 15 25 I 5 15 2515 15 2515 15 



Water vapor pressure difference 



Is 1525'! 



15 15 2515 15STr'^l525 



Sep Oct 



Nov 



5 15 25 15 15 2515 15 251 

Dec Jan Feb Mar 



5 15 

Apr May Jun 



25 
20 
15 
10 
5 
0 
-5 
-10 
-15 
-20 





Figure 1 — Photosynthesis in cembran pine, 
photosynthetically active radiation, air and soil 
termperature, and air humidity measured at the 
timberline (Stillberg, 1,980 m a.s.l., 1979-80). 



63 



FIELD CLIMATE CONTROL 

In the late sixties, the thermoelectrically climatized gas 
exchange chambers were developed by Koch and others 
(1968). Hasler and Blaser (1981) measured the first daily 
courses in young cembran pines on the east- and the 
north-facing slopes of an avalanche gully at the timberline 
at Stillberg. The two sites exemplify marked differences 
in microclimate (mean values of soil surface temperatures 
per growing season differ by more than 10 °C). These 
differences in microclimate also £iffect the photosynthetic 
rates of afforested cembran pines. Carbon dioxide gain, 
summarized over a season, of the trees on the north- 
facing slope was 34 percent less them for those on the 
east-facing slope (Turner and Streule 1983). 

Gas exchange measured in combination with meteoro- 
logical parameters was registered in a branch of an older 
cembran pine at Stillberg (1,980 m a.s.l.) throughout 
a whole winter season in 1979-80 (fig. 1). Havranek 
(1981) published an annual course of growth and a 
carbon dioxide balance, taken near the Patscherkofel 
laboratory. During summer the measurements were 
taken with climatized chambers. 

In the middle of the 1980's the so called "forest decline" 
had high priority in forest research in Europe. The main 
work in this field was done on other tree species than 
cembran pine. Nevertheless, some investigations were 
made on this species. Havranek (1987) took measure- 
ments on cembran pine at the timberline to investigate 
stress caused by air pollution in combination with climatic 
factors. Genys and Heggestad (1978) did not find a high 
susceptibility of cembran pines to ozone and sulphur di- 
oxide, but MinarCic and KubiCek (1991) suggested that 
dust could influence the stomata in needles of cembran 
pines and thus affect photosynthesis. Lutz and others 
(1988) emalyzed the photos)nithetic pigments of cembran 
pines growing at different altitudes, which thus are ex- 
posed to different ozone concentrations. Nebel and Matile 
(1992) investigated the cause of uncommon yellowing of 
the needles of cembran pines by analyzing photosynthe- 
sis, nutrient relations, longevity, and senescence of these 
declining needles. 

The newest measm*ements in cembran pine from the 
Stillberg experimental station are presented in these pro- 
ceedings (Koike and others). These are probably the first 
measvirements of photosynthesis in cembran pine at the 
timberline with an eirtificially elevated atmospheric 
concentration. A detailed publication giving these results 
is in preparation (Koike and others 1993). 

CONCLUSIONS 

At the beginning of the research on cembran pine, about 
150 years ago, scientists worked mainly at the forest 
level. With the development of sophisticated measure- 
ment instruments, branches, twigs, or even needles were 
the focus of interest. Rarely did anyone try to extrapolate 
these results from the small parts to a whole tree, a prob- 
lem which has still not been satisfactorily resolved. Prob- 
ably we should try to go back directly to the level of whole 
trees or even of stands, especially when we want to under- 
stand changes caused by external influences, such as air 



pollution, rising carbon dioxide, changing UV radiation, 
and many others. 

ACKNOWLEDGMENT 

I thank Mrs. M. J. Sieber for critical editing of the 
English version of my manuscript. 

REFERENCES 

Baig, M. N.; Tranquillini, W. 1980. The effect of wind and 
temperature on cuticular transpiration of Picea abies 
and Pinus cembra and their significance in desiccation 
damage at alpine treeline. Oecologia. 47: 252-256. 

Baig, M. N.; Tranquillini, W.; Havranek, W. M. 1974. 
Cuticulare Transpiration von Picea abies- und Pinus 
cembra-Zvfeigen aus verschiedener Seehohe imd ihre 
Bedeutung fiir die winterliche Austrocknung der 
Baiune an der alpinen Waldgrenze. Centredblatt fiir 
das gesamte Forstwesen. 91: 195-211. 

Bamberg, S.; Schwarz, W.; TranquiUini, W. 1967. Influen- 
ce of daylength on photosynthetic capacity of stone pine 
(Pinus cembra L.). Ecology. 48: 264-269. 

Caldwell, M. M. 1970. The effect of wind on stomatal 
aperture, photosynthesis, and transpiration of Rhodo- 
dendron ferrugineum L. and Pinus cembran L. 
Centralblatt fur das gesamte Forstwesen. 87: 193-201. 

Cartellieri, E. 1935. Jahresgang von osmotischem Wert, 
Transpiration und Assimilation einiger Ericaceen der 
alpinen Zwergstrauchheide und von Pinus cembran. 
Jahrbiicher fiir wissenschaftliche Boteuiik. 82: 460-506. 

Fankhauser, F. 1853. Uber das Vorkommen und 
Gedeihen der Arve. Schweizerisches Forst Journal. 
S83. 

Fankhauser, F. 1903. Der Kiefemschiittepilz an der Arve. 
Schweizerische Zeitschrifl fiir Forstwesen. 54: 321-323. 

Genys, J. B.; Heggestad, H. E. 1978. Susceptibility of dif- 
ferent species, clones and strains of pines to acute in- 
jury caused by ozone and sulfur dioxide. Plant Disease 
Reporter. 62(8): 687-691. 

Gtinthardt, M. S. ; Wanner, H. 1982. Die Menge des 
cuticularen Wachses auf Nadeln von Pinus cembran 
(L.) und Picea abies (L.) Karsten in Abhangigkeit von 
Nadelalter und Standort. Flora. 172: 125-137. 

Hasler, R.; Blaser, P. 1981. Nettophotosjmtheseaktivitat 
von Arve imd Larche auf verschiedenen Standorten der 
subalpinen Stufe. Mitteilungen der forstlichen 
Bundesversuchsanstalt, Wien. 140: 179-188. 

Havranek, W. 1972. The influence of soil temperature on 
photosynthesis and transpiration of young trees and on 
dry-matter production, at the timber line. Angewandte 
Botanik. 46: 101-116. 

Havranek, W. M. 1981. Stammatmung, Dickenwachstum 
und Photosynthese einer Zirbe {Pinus cembra) an der 
Waldgrenze. Mitteilimgen der forstlichen Bundes- 
versuchsanstalt, Wien. 142: 443-467. 

Havranek, W. M. 1987. Physiologische Reaktionen auf 
Klimastress bei Baumen an der Waldgrenze. Sympo- 
smm Klima und Witterung in ZusEunmenhang mit den 
neuartigen Waldschaden. Tagungsbeitrage, Miinchen: 
115-138. 



64 



Havranek, W. M.; Benecke, U. 1978. The influence of soil 
moistiire on water potential, transpiration and photo- 
synthesis of conifer seedlings. Plant and Soil. 49(1): 
91-103. 

Holzer, K. 1958. Die winterlichen Veranderungen der 
Assimilationszellen von Zirbe (Pinus cembra L. ) und 
Fichte (Picea excelsa Link. ) an der alpinen Waldgrenze. 
osterreichische botanische Zeitschrift. 105: 323-346. 

Keller, C. 1890. Tierische Forstbeschadigiingen an der 
Arve. Osterreichische Forstzeitung: 267-268. 

Keller, C. 1901. Die Arven-Erkrankungen im 
Oberengadin. Schweizerische Zeitschrift fur Forst- 
wesen. 52: 293-299. 

Keller, C. 1910. Die tierischen Feinde der Arve. 
Schweizerische Centralanstalt fur forstliches. Versuchs- 
wesen, Mitteilungen. 10(1): 1-50. 

Keller, T. 1970. Uber die Assimilation einer jungen Arve 
in Winterhalbjahr. Biindnerwald. 23(2): 49-54. 

Keller, T.; Beda-Puta, H. 1973. Gas exchange of leafless 
shoots in winter. Schweizerische Zeitschrift; fiir 
Forstwesen. 124(6): 433-441. 

Kemer, A. 1866. Wachstumsbedingungen der Zirbe. 
osterr. Monatsschr. Forstwesen. 1-15. 

Koch, W.; Klein, E.; Walz, H. 1968. Neuartige 
Gaswechsel-Messanlage fiir Pflanzen in Laboratorium 
und Freiland. Siemens Zeitschrift;. 42: 392-404. 

Koike, T.; Hasler, R. ; Matyssek, R.; Item, H. 1993. Sea- 
sonal changes in the photosynthetic capacity of Larix 
decidua and Pinus cembran planted on contrasting 
slopes at the timberline at Stillberg (Davos), eastern 
Switzerland. [In preparation]. 

Kronfiiss, H. 1980. Das Bestandesklima einer 
HochlagenaufForstung auf einem Sonnenhang. In: 
Beitrage zur subalpinen Waldforschung. 2. Folge. 
Mitteilvmgen der forstlichen Bundesversuchsanstalt, 
Wien. 129: 81-103. 

Kronfuss, H. 1983. Das Hohenwachstum einer Zirben- 
aufforstung auf einem Siidhang in klimatologischer 
Sicht. Allgemeine Forstzeitung. 94(12): 330-332. 

Kronfuss, H. 1986. Die Abhangigkeit des Hohen- 
zuwachses der Zirbe von der Seehohe. Jahresbericht 
1985. Forstliche Bundesversuchsanstalt, Wien: 137-144. 

Lutz, C; Schulte-Hostede, S.; Kirchner, M.; Reuther M. 
1988. Photosjnithetische Pigmente aus Nadelbaumen 
unterschiedlicher Hohenstufen des Otztals. Internatio- 
nales Symposium ''Verteilung und Wirkimg von Photo- 
oxidantien im Alpenraum. Sitzimgsberichte Miinchen 
1988: 415-425. 

MinarCic, P.; KubfCek, F. 1991. Localization of emissions 
on the needle surface. 1. Methodology, morphological 
characterization of the stomata {Picea abies, Pinus 
cembran). Ekologia, CSFR. 10(4): 405-413. 

Nebel, B.; Matile, Ph. 1992. Longevity and senescence 
of needles in Pinus cembran. Trees. 6(3): 156-161. 

Nevole, J. 1914. Die Verbreitimg der Zirbe in der osterr. - 
ungar. Monarchic. Verlag W. Frick Wien. 89 p. 

Pisek, A.; Winkler, E. 1958. Assimilationsvermogen und 
Respiration der Fichte (Picea excelsa Lind. ) in verschie- 
dener Hohenlage und der Zirbe (Pinus cembran L. ) an 
der alpinen Waldgrenze. Planta. 51: 518-543. 

Pisek, A.; Winkler, E. 1959. Licht- und Temperaturabhan- 



gigkeit der COg-Assimilation von Fichte (Picea excelsa 
Lind. ), Zirbe (Pinus cembran L. ) und Sonnenblume 
(Helianthus annuus L. ). Planta. 53: 532-550. 

Pisek, A.; Larcher, W.; Moser, W.; Pack, I. 1969. 
Kardinale Temperaturbereiche der Photosynthese 
und Grenztemperaturen des Lebens der Blatter ver- 
schiedener Spermatophyten. III. Temperatur- 
abhangigkeit und optimaler Temperaturbereich der 
Netto-Photosynthese. Flora B. 158(3): 608-630. 

Pisek, A.; Larcher, W.; Unterholzner, R. 1967. Kardinale 
Temperaturbereiche der Photosynthese und Grenz- 
temperaturen des Lebens der Blatter verschiedener 
Spermatophyten. I. Temperaturminimum der Netto- 
assimilation, Gefrier- und Frostschadenbereiche der 
Blatter. Flora B. 157(1): 239-264. 

Pisek, A.; Larcher, W.; Unterholzner, R.; Pack, I. 1968. 
Kardinale Temperaturbereiche der Photosynthese und 
Grenztemperaturen des Lebens der Blatter verschie- 
dener Spermatophyten. II. Temperaturmaximima der 
Netto-Photosynthese und Hitzeresistenz der Blatter. 
Flora B. 158(2): 110-128. 

Rikli, M. 1909. Die Arve in der Schweiz. Neue Denkschr. 
Schweizerische naturforschende Gesellschaft. 44. 455 p. 

Schonenberger, W.; Frey, W., eds. 1988. Untersuchungen 
zur Okologie imd Technik der Hochlagenauffbrstimg. 
Forschimgsergebnisse aus dem Lawinenanrissgebiet 
Stillberg. Schweizerische Zeitschrift fiir Forstwesen. 
139(9): 735-820. 

Schwarz, W. 1970a. Der Einfluss der Photoperiode auf 
das Austreiben, die Frostharte und die Hitzeresistenz 
von Zirben und Alpenrosen. Flora. 159: 258-285. 

Schwarz, W. R. 1970b. Frost hardiness, resistance to heat, 
and COg uptake in seedlings and mature stages of 
Pinus cembran L. Proc. 3. Forest Microclimate Sympo- 
sium, Kananaskis Forest Experimental Station; 1969; 
Seebe, AB: 137-145. 

Schwarz, W. 1971. The photos)nithetic capacity of some 
evergreens during the winter and the speed of their 
recovery of activity after sharp frosts. Berichte der 
deutschen botanischen Gesellschaft. 84(10): 585-594. 

Simony, F. 1870. Die Zirbe. Jahrbuch des osterreichischen 
Alpenvereins. 6: 349-359. 

Tranquillini, W. 1955. Die Bedeutimg des Lichtes und der 
Temperatur fiir die Kohlensaureassimilation von Pinus 
cembran-tJungwuchs an einem hochalpinen Standort. 
Planta. 46: 154-178. 

Tranquillini, W. 1957. Standortsklima, Wasserbilanz und 
COg- Gaswechsel junger Zirben (Pinus cembran) an der 
alpinen Waldgrenze. Planta. 49: 612-661. 

Tranquillini, W. 1958. Die Frostharte der Zirbe imter 
besonderer Beriicksichtigung autochoner imd aus 
Forstgarten stammender Jimgpflanzen. 
Forstwissenschaftliches Centralblatt. 77: 89-105. 

Tranquillini, W. 1959a. Die Stoffproduktion der Zirbe an 
der Waldgrenze wahrend eines Jahres. Planta. 54: 
107-129. 

TranquiUini, W. 1959b. Die Stoffproduktion der Zirbe 
(Pinus cembran L. ) an der Waldgrenze wahrend eines 
Jahres. II. Zuwachs imd COg-Bilanz. Planta. 54(2): 
130-151. 



65 



Tranqmllini, W. 1963a. Der Jahresgang der CO^- 
Assimilation junger Zirben. Mitteilungen der forst- 
lichen Bundesversuchsanstalt, Wien. 60: 501-534. 

Tranquillini, W. 1963b. Die COg-Jahresbilanz und die 
Stoifproduktion der Zirbe. Mitteilungen der forstlichen 
Biindesversuchsanstalt, Wien. 60: 535-546. 

Tranquillini, W. 1963c. Die Abhangigkeit der Kohlen- 
saureassimilation junger Larchen, Fichten und Zirben 
von der Luft- und Bodenfeuchte. Planta. 60: 70-94. 

Tranquillini, W. 1963d. Uber die Frostresistenz der Zirbe. 
Mitteilungen der forstlichen Bundesversuchsanstalt, 
Wien. 60: 547-562. 

Tranquillini, W. 1965. Uber den Zusammenhang zwischen 
Entwicklungszustand und Diirreresistenz junger Zirben 
{Pinus cembran L.) im Pflanzgarten. Mitteilungen der 
forstlichen Bundesversuchsanstalt, Wien. 66: 241-271. 

Tranquillini, W. 1979. Physiological ecology of the alpine 
timberline. Berlin: Springer- Verlag. 137 p. 



Tranquillini, W.; Machl-Ebner, I. 1971. Uber den Einfluss 
von Warme auf das Photosynthesevermogen der Zirbe 
(Pinus cembran L.) und der Alpenrose (Rhododendron 
ferrugineum L.) im Winter. Report of the Kevo Subarc- 
tic Research Station. 8: 158-166. 

TranquilUni, W.; Schiitz, W. 1970. Uber die Rinden- 
atmimg einiger Baume an der Waldgrenze. Central- 
blatt fiir das gesamte Forstwesen. 87: 42-60. 

Turner, H.; Streule, A. 1983. Wurzelwachstum iind 
Sprossentwicklung junger Koniferen im Klimastress 
der alpinen Waldgrenze, mit Berucksichtigung von 
Mikroklima, Photosynthese und Stoffproduktion. 
Irdning, Austria: Bundesanstalt Gimipenstein: 617-635. 

Ulmer, W. 1937. Uber den Jahresgang der Frostharte 
einiger immergriiner Arten der alpinen Stufe, sowie der 
Zirbe und Fichte. Jahrbiicher fur wissenschaftliche 
Botanik. 84: 553-592. 



66 



A PATTERN OF PINUS PUMILA SEED 
PRODUCTION ECOLOGY IN THE 
MOUNTAINS OF CENTRAL 
KAMTCHATKA 

Peter A, Khomentovsky 



Abstract— This paper discusses the results of Pinus pumila cone 
production studies in the mountains of central Kamtchatka. Pi- 
nus pumila was found to have a very high seed production poten- 
tial when not overtopped by neighboring trees. When shaded, 
few or no cones were produced on P. pumila. Seed production 
was not significantly dependent on altitude and site conditions. 
However, elevations above 1,300-1,400 m were unfavorable for 
cone survival, mainly because of low temperature and phenologi- 
cal delay. Successful seed production requires snow cover to 
protect ovulate buds on low-lying branches from freezing in the 
winter. In the summer, good seed production is related to high 
insolation, wind protection on cold sites, absence of shading, and 
good drainage for the root systems. At midelevations, under good 
seed production conditions in Kamtchatka, about 110 kg of seeds 
per hectare will be produced as compared to only 1 kg at the up- 
per elevational limits. Pinus pumila appears to have a high evo- 
lutionary potential that reveals itself in prominent and wide 
polymorphism of reproductive organs. 



The main purpose of this paper is to give a brief descrip- 
tion of some reproductive characteristics and ecological 
features of typical subalpine dwarf pine (Pinus pumila 
[Pall.] Regel) (Pinaceae, Strobus) (Pp) for many regions 
of its area ecotopes of the central Kamtchatka Mountains 
in northeastern Asia (fig. 1). 

Some years ago, we tried to identify the response of 
seed production (reflected in cones, seed, and nuclei masses 
and sizes) to some of the most evident differences of Pp 
environments, mainly to ecotopes, altitudinal position, 
and their location in regions with maritime or continental 
climate. We noticed (Khomentovsky and Khomentovskaya 
1990) that there is no strong dependence of seed param- 
eters and cone crop on the ecotope, elevation, and its posi- 
tion in the relief, Pp was supposed to have some ecotopic 
(macro- or microclimatic) optimum of seed production, 
which could be indirectly characterized by the climate 
within the belt of 600 to 900 m above sea level in the inte- 
rior continental climate regions of Kamtchatka peninsula, 
and widened in its lower limit almost to sea level near the 
coast. 

At the same time we tried to develop some of the first 
descriptions of conophagous insects feeding on Pp and to 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Peter A. Khomentovsky is Head of the Biogeocenology Laboratory, 
Kamtchatka Institute of Ecology and Nature Management, Prospect 
Rybakov, 19-A, Petropavlovsk-Kamtchatsky 683024 Russia. 



define their role in cone crop size and quality variation. 
We found (Khomentovsky and Efremova 1991) that the 
only two presently known insect species inhabiting Pp 
cones are Cecidomyia pumila Mamaev and Efi-emova 
(Diptera, Cecidomyidae) and Eupithecia abietaria Groeze 
(Lepidoptera, Geometridae) (the taxonomy of the last spe- 
cies has to be rechecked), and they affect crop quality very 
little. Therefore, this damage could be treated as negli- 
gible or at least not proven to be important. 

These preliminary conclusions forced us to proceed with 
a more detailed investigation of Pp seed production. Hav- 
ing no possibilities for obtaining direct microclimatic in- 
formation, we decided to define possible principal responses 
of Pp seed production to elevation and ecotopical features 
within a small creek basin with a high altitudinal gradient. 

THE RESEARCH AREA 

The research area is situated almost in the geographi- 
cal center of the Kamtchatka (56 °N., 158 °E.), in the 
Sredinnij Mountain Range system, the main system 
in the peninsula (fig. 1). According to the botanical- 
geographical regionalization scheme of L. F. Kimitsin 
(1963), this territory is included in the "mountain above 
alpine tundra shrubby region." According to oar scheme 
(Khomentovsky and others 1989), it lies in the contact 
area of "Central Kamtchatka plainland-submountain 
province of conifer-stonebirch forests" and "Sredinno- 
Western middle mountain stone birch-tundra forest prov- 
ince." The area covers the upper part of the conifer forest 
vegetation belt, subalpine, alpine, and above-alpine belts 
in Kamtchatka (figs. 1 and 2). 

Climate is subcontinental, with 300 to 400 mm of an- 
nual precipitation, moderately cold and snowy winters, 
and cool summers (monthly average air temperature in 
Esso village is 3 °C in May, 9.5 °C in June, 13 °C in July, 
12 °C in August, and 6 °C in September. The permanent 
snow cover at elevations of 700 m and higher begins to 
accumulate in the first half of October and disappears 
mEiinly in May. 

Orographically, the area can be called "typical middle 
altitudinal-mountainous" (according to common under- 
standing), which for Kamtchatka actually means middle 
and high mountains (due to the compression of its altitu- 
dinal vegetation belts). Average height of the mountains 
is about 1,500 m. Geological and geomorphological struc- 
ture of the area is determined by volcanogenic basalts and 
andesits (Ql, Q2, Q4), and by quatenary glacial, fluvio- 
glacial (Q3), alluvial, and proluvial (Q4) deposits. Glacial 
relief is conspicuous, especially in the upper parts of river 



67 



I 




Figure 1 — General description of the research area in Kamtchatka 
peninsula. Top left: principal scheme of vegetation belts: 1 — subalpine, 
2 — boreal forest, 3 — alpine, 4 — coastal and lowland. Top right: main 
mountain ridges and hydrological net. Bottom left: research area, within 
small basin. Dots are points where material was collected. 



more or less permanent (on large time scale) weak or 
moderate volcanic ashfalls impacts (mainly from the east- 
ern part of peninsula). 

Due to the geographical position, mountain relief, and 
severe climate of the area, vegetation is not very diverse 
and exhibits a mosaic-like spatial pattern. River valleys 
and lower parts of their tributaries are occupied by Popu- 
lus suaveolens, Chosenia arbutifolia, Larix cajanderi, 



and creek basins. Soils are typical for the Kamtchatka 
Mountain regions classification (I. A. Sokolov 1973), with 
the additions by N. V. Kazakov (personal communication) 
They include turf illuvio-humus, turf illuvio-volcanic ash, 
tundra illuvio-humus, tundra illuvio-humus- volcanic ash, 
and turf primitive. Soil genetics reveal themselves in the 
existence of specific features including the long-term fi'eez- 
ing period, anaerobic reactions in the lower horizons, and 



68 




Figure 2 — A fragment of the vegetation cover in the upper part of the altitudinal transect (see also fig. 1). 



Betula kamtschatica, andB. ermanii. Two dwarf tree spe- 
cies, Pinus pumila audi Alnus kamtschatica, are widely 
spread along the whole vegetation profile. They occur by 
groups or strips along the moraine tongues, river terraces, 
fluvioglacial deposits, and watershed ridges and slopes, 
and form the upper limit of woody vegetation (fig. 2). 

Plant formations in the middle and high altitudes are 
almost the same (except for Populus and Chosenia), but, 
moving upward along the narrowing valleys and water- 
sheds, along the altitudinal environmental gradient of 
gradually decreasing plant life conditions, we can see 
the distribution pattern of the formation changing from 
"macromosaic" to "micromosaic." Tree distribution is 
more strictly related to soil temperature conditions and 
drainage gradients, and to sites sheltered from wind and 
snow abrasion. This is true even for such ecologically 
flexible dwarf species as Pinus pumila and Alnus 
kamtschatica. 

The upper elevational limits of upright-grovnng woody 
vegetation {Larix kamtchatica, Betula ermanii) are lo- 
cated between 900 to 1,000 m, depending on site condi- 
tions. Pinus pumila and Alnus kamtschatica occur up to 
1,300 to 1,400 m (rare cl\mips). Up to 1,200 m P. pumila 
is able to produce seeds. 

MATERIALS AND METHODS 

Data were gathered mostly in September to October 1990 
(with some additional data from other years) at the field 
base "Bolgit," in the vicinity of Esso village (Bistrinsky 
district of Kamtchatka region), in the "Tupikin clyutch" 
creek basin, within an area of about 10 km^. This paper 
presents the material from one slope exposed to the east- 
northeast and from the area nearest to the watershed (up 
to 50 m from the slope edge), within an altitudinal range 
of 650 to 1,030 m. Permanent plots on an existing tran- 
sect and some additional temporary points were chosen 
for the cone crop measuring and sampling (fig. 1). 

The follovmig information was collected on each sample 
site (table 1): type of plant commimity, topographic 



location and ground cover of Pp clumps, average seasonal 
height of skeleton branches, general moisture conditions, 
shading by neighboring upright trees or climips. 

EXPLANATORY COMMENTS 

First, strictly speaking, we cannot consider Pp plants 
as common trees — they are not standing separately and 
do not have a single trunk; in most cases they overlap 
each other, both aboveground and in imderground parts. 
Also, we cannot call them "clones" because we do not know 
their origin and level of genetic relationships — which is 
especially important in the case of seed dispersal by ani- 
mals. The best term would be "clump" (we are grateful to 
Dr. Diana Tomback for some explanations on this topic). 

Second, the single trunk (in traditional understanding) 
of Pp is considered a creeping tree not a bush, and its 
length is only several centimeters or tens of centimeters. 
Branches of first and second order (I prefer to call them 
"trunk-branches") exhibit the same physiognomy, bear 
similar quantity of shoots, and form multi trunk-like 
crovms. In the case of syngenetic origin and continuous 
dispersal, Pp plants form such dense cover by overlapping 
basal parts in the litter and root systems that it is impos- 
sible to identify single plants and thus their total num- 
bers without destroying the stand. This forced us to use 
the only acceptable method of measuring cone crops or 
similar parameters — instead of measuring only one tree 
or clump, measure by square unit such as hectares or 
square meters. Certainly, the stand structure and charac- 
teristics of groimd cover have to be taken into account, 
especially for practical taxation. 

Third, it is also impossible, in many cases, to determine 
real age of adult Pp plants. They produce adventitious 
roots, are growing apically, and are gradually decaying 
in their basal parts during most of their lifetime, theoreti- 
cally endlessly. We have to recognize that we really can 
only know the age at the moment of measuring, similar 
to being counted in young trees, and not the real one, 
which is usually higher. 



69 



Table 1 — Short description of the ecotopes and Pinuspumila clumps being analyzed. The age of clumps in these analyses lies between 150 
and 260 years' 



Altitude, 




Pinus pumila 


Way of 
main water 


Woodstand 


Pinus pumila 


Shading 


m asl 


Position In the relief 


community type 


supply 


composition 


H avg, cm 


SO, % 


UTr 


Ppc 


1,030 


Upper part of the flat 
watershed, shaded from 
S and E by the ridge 


Pumilae pinetum 
carioso-hypnoso- 
ericosum, with 
fragments of P.p. 
cladinosum (CHE*C^ 


A+S 


lOPp 


100 (15-20) 


80 unev 


0 


1 


950 


Plateau, opened to all 
directions, watershed 
with slight slope to 
NNE 


P.p. purum, with 
fragments of 
P.p. carioso- 
cladinosum (P*CC^ 


A 


lOPp 


40 (10-15) 


40 unev 


0 


0 


900 


Flat watershed above 
the creek source, 
shaded by the ridge 
from S 


P.p. hypnoso- 

carioso-ericosum 

(HCE) 


A+S 


lOPp+Lk 
(upper 
limit of 
Lk) 


300 


40 unev 


1 


1 


810 


A ridge of the complex 
watershed, slight 
slope to the NE 


P.p. carioso- 

hypnoso-ericosum 

(CHE) 


A 


lOPp+Lk 


150 (35-40) 


60 unev 


1 


2 


800 


Middle part of the E 
exposed slope of wide 
creek valley 


P.p. hypnoso- 

carioso-ericosum 

(HE) 


S+A 


lOPp+Lk 


200 


80 


0 


1 


680 


Eastern border of the 
watershed ridge with 
flattened top 


P.p. hypnoso- 

carioso-ericosum 

(HCE) 


A+S 


7Pp 3Lk 


300 


60 unev 


2 


1 


650 


Lower part of E 
exposed slope in 
narrow creek valley 


P.p. ericoso- 
sphagnosum 


8 


lOPp+Ak 


300 (40-45) 


100 


0 


3 



^ Pinus pumila community type: in parentheses — abbreviation for fig. 8; Way of main water supply: A — atmospheric, S — slope; Woodstand composition: Pp — Pinus 
pumila, Lk — Larix cajanderi, Ak — Alnus kamtschatica (dwarf alder); Pinus pumila: H avg — average height (cm), in parentheses — rough estimation of annual shoot 
elongation (mm); SC% — surface covering by the clumps, "unev" — uneven; Shading: Utr— from upright trees of different species, Ppc — from Pinuspumila neighbor- 
ing clumps; 0 — no shading, 1 — slight, 2 — moderate, 3 — heavy shading. 



In each point for data and sample collecting v^^e used 
3 to 5 small randomized plots of different size — from 1 to 
15 m^. The following parameters were countered in place 
(table 2): number of skeleton branches ("trunk-branches"), 
number of germinating shoots on each skeleton branch; 
number of current year crop cones (the second year of 
cone development) and number of next year crop cones 
(female buds of the first year) on each germinating shoot 
and each skeleton branch; and among the total quantity of 
each year's cones the number of cones damaged by insects 
iCecidomyia pumila — damage in the first year of cone de- 
velopment, Eupithecia abietaria — damage in the second 
year of cone development), and by the nutcracker {Nuci- 
fraga caryocatactes kamtschatkensis) . 

At each site, 10 to 100 cones of the current year's crop 
were collected and measured before they dried (cone 
length and cone diameter). Later, after air drying during 
some weeks or months, cones and seeds were measured 
by 16 other parameters, six of which are discussed in this 
paper: cone mass; 1,000 seeds mass; 1,000 nuclei mass 
(both by measuring a number of samples with 50 seeds in 
each); nimiber of seed scales in the cone; among the total 



number of seed scales, the number of scales not contain- 
ing seeds under them; and total seed quantity in the cone. 

RESULTS AND BRIEF COMMENTS 

• The structure of the cone-bearing portion of the 
crown and cone crop estimates (tables 1 and 2, figs. 1-6): 

As a whole, the number of skeleton branches and germi- 
nating shoots per hectare varied very little within alti- 
tudes of 650-800 m and definitely increased upward. This 
could be explained by increasing isolation. Most likely it 
was the reason for the especially increasing number of 
germinating shoots (two to three times higher than skel- 
eton branches) and, correspondingly, the nvimbers of 
current-year and next-year cones. The last varied more 
than the first because some of them vdll inevitably die in 
the natural selection process during the maturation time 
in the coming year. 

The number of shoots with cones per skeleton branch 
(fig. 4) changed opposite to that of the number of skeleton 
branches per 1 hectare (we noticed the increased number 



70 



of shoots, including nongerminating). Even a slight plant 
shading immediately resulted in a change in the number 
of shoots (sites 900 m and 1,030 m). 

The quantity of current and next-year crop cones per 
one germinating shoot had specific dispersal features: 
at each site, especially in the lower, shaded part of the 
elevation profile, cones of any one year were overwhelm- 
ing. This migration of intensive seed-producing centers 
from year to year supports our previous conclusion 



(Khomentovsky and Efremova 1991) about the existence 
of local seed production mosaics, which provides a con- 
tinuous supply of seeds, important not only for guaran- 
teed reproduction and microevolutionary diversity for Pp, 
but also for all zooconsumers such as nutcracker, other 
birds, and mammals. 

• Cones and seeds mass and size variation (tables 1, 3; 
figs. 7 and 8): 



Table 2 — Characteristics of germinating crown parts and cone crop outcome of Pinus pumila in central Kamtcliatka Mountains^ 



nn 


Parameter 






Altitude (m above sea level) 






650 


680 


800 


810 


900 


950 


1,030 


1 


No. of skeleton branches/ha 


1,200 


4,167 


2,240 


21,111 


11,000 


30,000 


10,333 


2 


No. of germinating shoots/ha 


1,733 


7,917 


3,200 


47,778 


57,000 


86,667 


33,000 


3 


No. of this year crop cones/ha 


1,733 


4,375 


3,136 


2,389 


46,170 


56,334 


28,000 


4 


No. of next year crop cones/ha 


260 


0 


160 


51,122 


19,950 


143,000 


24,550 


5 


No of aerminatina shoots/skel bran 


1.44 


1.90 


1.43 


2.26 


5.19 


2.89 


3.00 


6 


No of thi^ vppir oroo conp^/<>kpl hri^n 


1.44 


2.00 


1.40 


0.11 


4.20 


1.88 


2.56 


7 


No of next vear croo cones /skel bran 


0.22 


0 


A A7 


O AO 


1 Md. 


4.77 


2.54 


o 
O 


No. of this year crop cones/germ, shoot 


1 .00 


H AC 
1 .00 


0.98 


0.05 


0.81 


A 


A 7ft 


9 


No. of next year crop cones/germ, shoot 


0.15 


0 


0.05 


1.07 


0.35 


1.65 


0.86 


10 


Percentage of skeleton branches with 


















various numbers of germinating shoots: 


















with one shoot 


67 


30 


64 


58 


28 


22 


32 




with two shoots 


22 


CA 

50 


29 


16 


18 


J4 






with thrfifi shoots 

will 1 11 II Ol Iwwlw 


11 


20 


7 


16 


0 


22 


8 




with four shoots 


0 


0 


0 


0 


9 


11 


4 




with fiv/o chnr^tQ 
Will i live? Ol IwUlO 


0 


0 


0 


5 


9 


0 


8 




with qIy QhnntQ 

Willi OlA Ol lUUlO 


0 


0 


A 
0 


A 
0 


A 
0 


0 


2 




with seven shoots 


0 


0 


0 


0 


0 


0 


5 




with eight shoots 


0 


0 


0 


0 


9 


11 


5 




with nine shoots 


0 


0 


0 


0 


g 


0 


0 




with ten shoots 


0 


0 


0 


0 


0 


0 


2 




with eleven shoots 


0 


0 


0 


0 


0 


0 


0 




with twelve shoots 


0 


0 


0 


5 


0 


0 


0 




with thirteen shoots 


0 


0 


0 


0 


Q 


0 


1 


1 1 


r ciccf iictgc ui gcniiiriaurig biiouis Dcaiirig 


















various number of this year cones (T) 


















and next year cones (N): 


















0N-1T 


81 


100 


93 


86 


63 


8 


32 




ON -21 


7 


0 


2 


8 


4 


0 


6 




Cumulative percent of shoots with 


















this year's crop cones only 


88 


100 


95 


94 


67 


8 


38 




1N-0T 


4 


0 


0 


3 


25 


15 


30 




2N-0T 


4 


0 


5 


0 


2 


11 


7 




3N-0T 


0 


0 


0 


0 


0 


11 


0 




Cumulative percent of shoots with 


















this year's crop cones only 


8 


0 


5 


3 


27 


37 


37 




1N-1T 


4 


0 


0 


3 


3 


23 


10 




1N -2T 


0 


0 


0 


0 


3 


4 


2 




2N - 1T 


0 


0 


0 


0 


0 


15 


6 




2N-2T 


0 


0 


0 


0 


0 


5 


5 




3N-1T 


0 


0 


0 


0 


0 


8 


1 




4N - 1T 


0 


0 


0 


0 


0 


0 


1 




Cumulative percent of shoots with 


















cones of both years 


4 


0 


0 


3 


6 


55 


25 



'Only the area covered with P'mus pumila clumps is taken into account (par. 1 -4), not the whole territory of the site; number of next-year cones can decrease 
during the year due to natural selection processes. 



71 



Table 3 — Some characteristics of Pinus pumila cones and seeds in various elevations in central 
Kamtchatka Mountains (within research area)' 



Regression on altitude Correlation Average value 



CL 


= 54.71 





0.020 A 


R 


= -0.73 


AVG (cl) = 


43 


+/- 


8 (25-62) 


CD 


= 33.68 


- 


0.010 A 


R 


= -0.79 


AVG (cd) = 


27 


+/- 


2(18-37) 


CM 


= 7.81 




0.030 A 


R 


= -0.41 


AVG (cm) = 


7 


+/- 


1 (4-10) 


SSN 


= 32.30 


+ 


0.010 A 


R 


= 0.53 


AVG(ssn) = 


39 


+/- 


5 (26-52) 


SWS 


= 9.06 


+ 


0.004 A 


R 


= 0.33 


AVG (SWS) = 


11 


+/- 


3 (5-18) 


SQT 


= 28.50 


+ 


0.003 A 


R 


= 0.76 


AVG (sqt) = 


45 


+/- 


5 (26-68) 


SMP 


= 46.71 


+ 


0.002 A 


R 


= -0.11 


AVG(smp) = 


45 


+/- 


3 (32-55) 


SMTH 


= 101.67 




0.020 A 


R 


= -0.45 


AVG (smth) = 


84 


+/- 


8 (52-116) 


NMTH 


= 56.49 




0.020 A 


R 


= -0.49 


AVG (nmth) = 


43 


+/- 


4 (27-68) 



'CL — cone length, mm; CD — cone diameter, mm; CM — cone mass, g; A — altitude, m above sea level; SSN — seed 
scale number; SWS— number of scales without seeds (within SSN); SQT — total number of seeds in the cone; SMP — 
percentage of seeds mass in cone mass; SMTH — mass of 1 ,000 seeds, g; NMTH — mass of 1 ,000 nuclei, g. In paren- 
theses at the right side — minimum and maximum values. 



Mass variation (here and later we speak about current- 
year crop cones only) is more evident in the upper part of 
the elevation profile, in micromosaic sites with increasing 
abiotic environmental pressure. In sufficiently insolated 
(sunny) and wind-protected sites cone mass is increasing, 
in shaded sites or windy plateaus — decreasing. At the up- 
per elevation levels of Pp distribution (1,200-1,300 m) we 
observed natural abrupt decreases of all seed-producing 
parameters. It is remarkable that in general cones and 
seeds masses are not too closely related to definite altitu- 
dinal or site positions of the tree, even in cases of evi- 
dently unfavorable envirormients (site 950 m). 

Self regulation of seed production is obvious also: along 
the elevation gradient, number of seed scales in the cone 



(with and without seeds) and total number of seeds in the 
cone slightly increases as well, and at the same time seed 
and nuclei mass is unchanging or slightly decreasing. It 
means that seeds become smaller but grow in quantity, 
keeping the same reproduction potential and having more 
protection from environmental impacts. 

Cone size variation, most evident in their length, corre- 
sponds vrith mass variation in relation to site position in 
the landscape. Here, as well as in the previous case, the 
main impact factor is not elevation but landscape struc- 
ture of the ecotope (slopes in wide or narrow valley, pla- 
teau, watershed ridge, etc.). Biocenotic structure is not 
too important in this case, at least at the forest type level. 
All ecotopes, excluding the 950-m site (opened plateau), 



160 



140 



§ 120 
O 



£ 100 

(0 

u 

X 80 

CL 

O 60 

n 
E 

3 

Z 40 



20 





























i 


\ 

\ 














i 
i 


\ 

\ 

— \ 














— i-/- 


\ 

. \ 














/ 
/I 


\ \ 
\\ 

V 












\ 1 


• y 












>' 
/ 

f 


\l 
1 • 


1 . 












1 / 


1 ^ 









skeleton branches 

germinating shoots 

— current year cones 
next year cones 



650 680 800 810 900 950 1.030 

Altitude (m as!) 

Figure 3 — Principal structure of P'mus pumila crowns at 
various altitudes and estimated cone crop per hectare. 



72 



6 



germ, shoots/sk. br. 




650 680 800 810 900 950 1,030 



Altitude (m asl) 

Figure 4 — Structure of Pinus pumila crown germinating 
part (number of germinating shoots and estimated cone 
crop per skeleton branch). 



are occupied by the same group of forest types — Pumilae 
pinetum carioso-hypnoso-ericosum, and its variations. 
Generally speaking, cones size, as well as cones mass, 
which has good correlation with them (r = 0.7-0.8), vary 
within limits already known for the species. 

• Damage caused by insects and birds (fig. 9): 

This was investigated in some detail previously 
(Khomentovsky and Efremova 1991) and is not likely 



to effect any substantial change in average cone sizes: it 
does not affect cone diameter and leads to cone length de- 
creases up to 3-9 percent (Cecidomyia pumila). However, 
in some cases this insect species damage results in under- 
development of up to 20 percent of the seeds, but this im- 
pact usually does not stop the development of the rest of 
the seeds in the cone and is compensated at the popula- 
tion level by the abundant production of seeds. 




73 



100 




cones of both years 
next year cones only 
current year cones only 



650 680 800 810 900 950 1,030 

Altitude (m asl) 

Figure 6 — Distribution of current-year and next-year cone crop on 
germinating shoots. 



The level of damage caused by Cecidomyia pumila, 
clearly noticeable visually by the cone curving, is usually 
rather high everywhere but it varies with different site, 
environmental, and weather conditions, to which insects 
are more sensitive. In our case it was changing from mod- 
erate damage of 25-50 percent on the 650-m site (slope 
in the narrow valley), the 800-m site (slope in the wide 



valley), and the 900-m site (wind-protected watershed at 
the upper limit of Larix) to a maximum of 100 percent on 
the 810-m site (watershed ridge) and a minimal 8 percent 
on the 1,030-m site (weakly wind protected and too cold 
for insects ecotope). 

Confirmation of the negligible character and amount of 
insect damage is also seen in the level of seed extraction 



o 



o 
m 

w 

"O 

<u 
a> 
0) 
o 
in 

V) 

<a 
c 
o 
O 




650 



800 



900 



1,030 



O 

o 

CM 
II 

o 

(/) 



Altitude (m asl) 



cone mass 
SM std 



CM std 

50 nuclei mass 



50 seeds mass 
NMstd 



Figure 7 — Variations in Pinus purpila cones, seeds, and nuclei mass. 
SL — slopes, P — Plateau. 



74 




650 680 800 810 900 950 1,030 



Altitude (m asl) 

Figure 8 — Variations in Pinus pumila cone size. 1 — forest type (see table 1). 



by the nutcracker, which collects only full, matured seeds 
(fig. 9). 

DISCUSSION AND CONCLUSIONS 

Data analysis results, combined with some previous 
material, allow us to make some conclusions about Pinus 
pumila seed production in Kamtchatka, keeping in mind 
the fragmentary nature of our study, even though it was 
done in typical environmental conditions. 

1. The main conclusion is that P. pumila seed produc- 
tion has such a high potential in time and space that we 
can speak about the environmental (abiotic and natural 
biotic) impact on it in the frame of modification only but 
not in the frame of regulation. Environmental stresses 
usually cannot cause a crop crisis — trees have compensat- 
ing mechanisms both at the organismic and population 



level, and one of them is the microhabitat mosaics of an- 
nual seed production. 

2. Pinus pumila seed production is not (in Kamtchatka, 
at least) significantly dependent on altitude and site con- 
ditions. Landscape conditions (above facial level) are the 
most important. However, this is true only when trees 
are not overshadowed by neighboring upright trees; that 
is, when Pp forms an independent vegetational belt in 
more or less solid segments — in plain or mountain tundra, 
upper forest-tundra ecotone, on the seashore dunes, in 
wide river valleys, etc. Pp often grows as the lower canopy 
of more or less dense Larix forests (more than 30 percent 
of Larix cover), which have the same age and the same 
syngenetic origin (very often in pyrogenetic ecotopes). 
These shaded dwarf pine usually do not produce seeds 
and develop vegetatively only. 

3. The utmost upper limits of distribution (1,300-1,400 m 
above sea level) are unfavorable for Pp seed production. 




650 



680 



800 810 900 

Altitude (m asl) 



950 




1,030 



Ea+Nc 



□ 



Cp 



Ea+Cp+Nc 



Nc 



Figure 9 — Percentage of cones damaged by insects and birds. Cp — Cecidomyia 
pumila; Ea — Eupithecia abietaria; Nc — Nucifraga caryocatactes. 



75 



mainly because of low temperature and phenological de- 
lay. But even here, some cones can be found, and some- 
times the crop can be abundant in wind-protected habi- 
tats. Phenological delay often prevents the nutcracker 
from collecting seeds here. These seeds, having been dis- 
persed by birds from lower sites, and having more stable 
genotypes, provide a better chance for successful refores- 
tation in new areas. At the same time, dwarf pine regen- 
eration at high altitudes is more or less provided by the 
remaining abundant seeds. Besides this, numerous zoo- 
consorts (mainly micromammals) will have a good food 
supply most valuable in extreme conditions. 

The same unfavorable conditions for seed production 
can be found in the Pp belt at the seashore (Pacific 
Kamtchatka coast), on the dunes nearest to the water. 
But here, as well as in the mountains, trees utilize each 
opporttmity for maximal realization of their reproduction 
ability. Pp seed production on the second to third dune 
of the coast belt is equal in cone size and mass parameters 
to those at the 900- to 1,000-m altitude and in seed and 
nuclei parameters to those at the 650- to 750-m altitude 
in the continental climate portion of the peninsula. 

Conditions moderately favorable for Pp seed production 
can be found in plain and submoimtain (foothill) places, 
for instance on gravel-sandy fluvioglacial or proluvial 
deposits, where abundant cone crops sporadically occur. 

The most productive seed zone lies in the vegetation 
belt within the 600-900 m above sea level area. Here, 
cone crops are mostly permanent and stable from year 
to year. 

4. Successful seed production of Pp needs the following 
conditions: in winter — sufficient snow cover to protect buds 
of low-ljdng trees from freezing; in summer — high insola- 
tion, good wind protection on cold sites, absence of shad- 
ing, and good drainage for the root systems. The main 
adaptive feature of Pp is its ability to lay close to the 
ground in the autumn. This process is initiated by low 
temperature in the autumn before the snow cover. This 
unique feature provides good chances of survival under 
very extreme conditions. Strict dependence of Pp on snow 
cover depth is proved by coincidence of its area border with 
the snow depth isoline of 40 cm and more (Lukitcheva 
1964; Tstcherbakova 1964). 

5. Special research and preliminary work, including 
mapping of various productivity sites, is needed for the 
practical (silvicultural) estimation of current or future 
cone crops of Pp. It is possible to show, as an example, 
the scope of crop diversity. In the middle elevation site 
(700-800 m), with 5 g of seed mass in the cone, 80 percent 
of germinating shoots, and 80 percent of Pp covering the 
ground, we can get about 110 kg of seeds per hectare. At 
the upper limit of seed production, with 2 g of seed mass 
in the cone, 30 percent of germinating shoots, 10 percent 
of covering, we can get about 1 kg of seeds per hectare. 
High variability of cone crop is illustrated in table 1 data: 
taking into account only pure stands of Pp (without con- 
sidering its real distribution on the surface) we can see 
that the crop size on the 650-m site is about 6.5 kg/ha, 
and about 212 kg/ha on the 950-m site. 

6. Starting from the elevation that corresponds to the 
upper limit of upright trees distribution (900 to 1,000 m) 



in Kamtchatka, Pp produces an increasing number of 
shoots. This can be interpreted as a organismic compen- 
sation for the increasing severity of abiotic environmental 
conditions. Photosynthetic ability increases, seed produc- 
tion potential, as we saw, remains at the same high level, 
and seed protection (thickness of seed walls) becomes 
stronger. All this takes place mostly in the subalpine belt 
and proves Pp belongs mainly to this type of vegetation. 
Coincidence of these activated processes with the upper 
limit of upright trees distribution allows us to hypothe- 
size the existence of some kind of temperature threshold- 
trigger. It could be the sum of effective temperatures, 
above which the development of upright tree forms is pos- 
sible, and below which only prostrate forms develop. 

7. We agree with the opinion that Pp, as well as other 
stone pines, has a high evolutionary potential, revealing 
itself in prominent and wide polymorphism of reproduc- 
tive organs and their functional flexibility (Pravdin and 
Iroshnikov 1982). This is well illustrated by our work 
and other sources. Ecotypical features are also evident, 
accompanied by stability of the main seed-production 
parameters. In fact almost all of this is similar to that 
which is known from other Pp regions (Bobrinev and 
Rylkov 1984; Efremova and IvHev 1972; Kapper 1954; 
Khomentovsky and Khomentovskaya 1990; Krylov and 
others 1983; Rush 1974). 

8. Pp seed production potential exceeds the possibilities 
of its realization in our research area, and most likely, in 
the Kamtchatka peninsula as a whole. Ecotopical diver- 
sity in Kamtchatka is evidently only a part of the whole 
scope of acceptable conditions for the species (here we 
can admit that low temperatures affect vegetative growth 
three to four times more than they affect generative or- 
gans variability). Therefore, we should look for possible 
phylogenetic roots of this species in the regions with a 
wider geoecological spectrum of sites. 

9. Our observations (partly presented in this paper) al- 
low us to speculate that Pp is a middle-altitudinal species, 
with more preferable growth conditions in the moderately 
continental climate (in relatively humid mountains) than 
at the seashore. But, regarding the whole area, the defi- 
nite and necessary influence of Pacific wet air masses re- 
sults in considerable snow cover. 

Combining this conclusion with some known hypothe- 
ses of Pp evolution (Sotchava and Lukitcheva 1953; 
Tikhomirov 1949; etc.) we think that this dwarf tree spe- 
cies most likely appeared in Tertiary times in some cli- 
matically subcontinental mountain regions of Angarida 
(northeastern Asia). In Kamtchatka, pollen data show 
Pp has been here for 1-1.5 million years (Malaeva 1967; 
Tchelebaeva and others 1974). During Pleistocene inter- 
glacial periods, and in the early middle Holocene, Pp 
occupied periglacial zones and other places unacceptable 
for more-thermophil tree species. The highest incidence 
of polyembriony and underdevelopment of seeds in 
Kamtchatka support the possible centers of origin in the 
moderate continental climates when compared to Magadan 
(northeastern continental part of Asia) and Buryatia 
(southern part of eastern Siberia) (Iroshnikov 1972; Rush 
1974). These features can be considered a reaction to less 
favorable conditions of the secondary environments. 



76 



including Kamtchatka peninsula. According to A. I. 
Iroshnikov (1972), Pp is one of the most well-adapted 
species to the continental climate of northeastern Asia. 

REFERENCES 

Bobrinev, V. P.; Rylkov, V. F. 1984. Seed production of 
dwarf stone pine. Inform, bull, of Tchitinsky Informa- 
tion Center, #76-84. Tchita. 5 p. [In Russian]. 

Efremova, L. S.; Ivliev, L. A. 1972. On dwarf stone pine 
seed producing at Kamtchatka. In: Utilization and re- 
producing of forest resources in the Far East. Abstr. of 
the all-imion conference Khabarovsk. Part 2: 158-159. 
[In Russian]. 

Iroshnikov, A. I. 1972. Seed producing and seed quality 
of conifers in northern and mountainous regions of 
Siberia. In: Seed producing of Siberian conifers (collec- 
tion of papers). Novosibirsk: 98-117. [In Russian]. 

Kapper, O. A. 1954. Conifers (forestry characteristics). 
IVloscow-Leningrad. 304 p. [In Russian]. 

Khomentovsky, P. A.; Efremova, L. S. 1991. Seed prod- 
uction and cone-feeding insects oiPinus pumila at 
Kamtchatka peninsula: aspects of coexistence. In: 
Baranchikov, Y. N.; Mattson, W. J.; Hain, F. P.; Payne, 
T. L., eds. Forest insects guilds: patterns of interaction 
with host trees. U.S. Department of Agriculture, Forest 
Service, Gen. Tech. Rep. NE-153: 316-320. 

Khomentovsky, P. A.; Kazakov, N. V.; Tchernyagina, O. A. 
1989. Tundra-forest zone at Kamtchatka: problems of 
conservation and management. In: Problems of nature 
management in taiga zone (collection of papers). 
Irkutsk: 30-46. [In Russian]. 

Khomentovsky, P. A.; Khomentovskaya, I. G. 1990. Geo- 
graphical variability of Pinus pumila seed producing at 
Kamchatka. In: Questions of Kamchatka geography. 
Petropavlovsk-Kamchatsky. No. 10: 47-55. [In Russian]. 



Krylov, G. V.; Talantsev, N. K; Kozakova, N. F. 1983. 
Siberian pine (Pinus sibirica Du Tour). Moscow. 216 p. 
[In Russian]. 

Kunitsin, L. F. 1963. A pattern of Kamtchatka nature 
regionalization. In: Nature conditions and regionaliza- 
tion of Kamtchatka district. Moscow: 7-26. [In Russian]. 

Lukitcheva, A. N. 1964. Distribution areas of Asia plants 
(a map). In: World geographical atlas. Moscow. 112 p. 

Malaeva, E. M. 1967. Kamtchatka vegetation develop- 
ment in Pliocene-Pleistocene. In: Development of Sibe- 
rian and far-eastern vegetation in Quaternary (collec- 
tion of papers). Moscow: 78-170. [In Russian]. 

Pravdin, L. F.; Iroshnikov, A. I. 1982. Genetics of Pinus 
sibirica Du Tour, P. koraiensis Sieb. et Zucc. and P. 
pumila Regel. Annales Forestales. 9(3): 79-123. 

Rush, V. A. 1974. Biochemical characteristics of five- 
needle pines seeds. In: Biology of conifer seed repro- 
duction in western Siberia (collection of papers). 
Novosibirsk: 180-184. [In Russian]. 

Sokolov, I. A. 1973. Volcanism and soils formation 
(with Kamtchatka as an example). Moscow. 224 p. 
[In Russian]. 

Sotchava, V. B.; Lukitcheva, A. N. 1953. On geography 
of dwarf stone pine. Reports of the USSR Academy of 
Sciences. 90(6): 1163-1166. [In Russian]. 

Tchelebaeva, A. I.; Shantser, A. E.; Egorova, I. A.; Lupikina, E. 
G. 1974. Cenozoic deposits of the Kuril-Kamtchatka 
area. In: Kamtchatka, Kuril and Commander islands 
(series of History of the relief development.) Moscow: 
31-57. [In Russian]. 

Tikhomirov, B. A. 1949. Dwarf stone pine {Pinus pumila 
(Pall.) Rgl), its biology and utilization. MOIP series, 
botan. division, Moscow. 14(6): 106 p. [In Russian]. 

Tstcherbakova, E. J. 1964. An average of maximal decade 
snow depths (a map). In: World geographical atlas. 
Moscow. 220 p. 



77 



NEEDLE LONGEVITY AND 
PHOTOSYNTHETIC PERFORMANCE IN 
CEMBRAN PINE AND NORWAY SPRUCE 
GROWING ON THE NORTH- AND EAST- 
FACING SLOPES AT THE TIMBERLINE 
OF STILLBERG IN THE SWISS ALPS 

Takayoshi Koike 
Rudolf Hasler 
Hans Item 



Abstract — The number of shoots in each age class and the pho- 
tosynthesis of the shoots were measured on cembran pine {Pinus 
cembra) and Norway spruce (Picea abies) growing on north- and 
east-facing slopes at 2,185 m above sea level. Needle longevity 
on the north-facing slope exceeded needle longevity on the east- 
facing slope by 1 to 3 years in both species. A negative correla- 
tion between maximum photosynthesis and needle lifespan was 
foimd. However, there was no physiological difference in the 
photosynthesis of trees on the respective slopes. 



In the Swiss Alps, cembran pine (Pinus cembra) grov^rs 
around timberline level and Nonvay spruce (Picea abies) 
from valleys to mountainsides (Kuoch and Amiet 1970). 
At the subalpine timberline, the growth and biomass in- 
crease of trees are strongly influenced by the microsite 
performance and microclimate (Livingston and Black 
1987; Schonenberger and Frey 1988; Tvu-ner and others 
1982). Using the monitoring experiment on photosyn- 
thetic production as a basis, the biomass increase of 
mountain pine (Pinus mugo) growdng at the timberline 
was found to be regulated by the soil temperature and the 
net radiation (Hasler 1982). However, physiological char- 
acteristics of trees were not revealed. Is the turnover rate 
of needles related to the tree size? What about the physi- 
ological adaption of needles to the microclimate? 

In this study, these questions, needle lifespan and the 
age-related photosynthesis of cembran pine planted on 
north- and east-facing slopes, were measured. We present 
the physiological parameters of needle photosynthesis. 
Anatomical traits of needles were investigated with refer- 
ence to the estimation of the intercellular CO^ concentra- 
tions (C.) (Terashima 1992). We discuss the trade-off rela- 
tion between the needle lifespan and photosynthetic rates 
influenced by the microclimate. 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Takayoshi Koike, Rudolf Hasler, and Hans Item are members of the 
"Ecophysiology" group, Swiss Federal Institute for Forest, Snow and Land- 
scape Research, CH-8903 Birmensdorf, Switzerland. Present address of 
T. Koike: Forest Products Research Institute, Sapporo 062, Japan. 



MATERIALS AND METHODS 

At the study site (Stillberg in the Swiss Alps), on the 
north-facing slope, the yearly net radiation surplus was 
308 kWh/m^ lower, and the root-zone temperature during 
grov^rth period was 2-4 °C lower than on the east-facing 
slope at 2,185 m above sea level (a.s.l.) (Turner and others 
1982). The mean soil temperature at 10 cm on the north- 
and east-facing slopes from June to September was 3.5 
and 4.5 °C, respectively (Koike and others, submitted). 

Plant Materials 

Plants studied were cembran pine, (provenance 2,050 m 
a.s.l.) and Norway spruce (provensmce 1,960 m a.s.l.), 
raised from seeds in the nursery at Birmensdorf ZH (550 m 
a.s.l.). The cembran pines were transplanted in 1975, 
and the Norway spruces in 1962, to the experimental site 
(2,185 m a.s.l.). Mean tree height as of 1988 on the north- 
facing slope and east-facing slope was 40 cm and 80 cm in 
cembran pine and 36 cm and 45 cm in Norway spruce, re- 
spectively. For measurements of photos3mthesis, attached 
shoots vdthin the even-aged needles were used. Younger 
shoots were removed when the photosynthesis of older 
shoots was measured. The cut shoot ends were sealed 
with vaseline. 

Measurements 

The gas exchange rate was measured with a thermo- 
electrically controlled chamber (Mini-cuvette Walz, FRG), 
which was coated with aluminium foil in order to fully 
use diffuse light (Koike and others, submitted). Photosyn- 
thesis was determined v^dth two infrared gas analyzers 
(Binos, Leybold-Heraeus, FRG). Dew point mirrors (Walz, 
FRG) were used to measure the absolute air hmnidity. 
Needle temperatures were monitored by a 0.1 mm chromel- 
constantan thermocouple. The PPFD above the needles 
was measured with a GaAsp photodiode (Hamamatsu 
Gil 18, J) after calibration vsrith a LiCor quanttmi sensor 
(USA). A COj dispensing apparatus (Walz, FRG) pro- 
duced different COg concentrations in the chamber. 



78 




Figure 1 — Cross section of a needle of a cembran 
pine (right) and a NoPA'ay spruce (left). 



Anatomy of needles was observed with a microscope 
after needles were sectioned with a microtome (Leitz). 
Total needle surface area of the used twig was estimated 
by the modified glass bead method (Davies and Benecke 
1980). The number of specimens for the census of needle 
longevity was more than 700 shoots of 10 trees. Dry 
weight of the needles was determined after 48 hours dry- 
ing at 80 °C. Nitrogen concentration in the needles was 
analyzed with a CHN analyzer (Rapid, USA). 

RESULTS 

Needle Characteristics 

The anatomy of needles of cembran pine and Norway 
spruce was similar to the "homobaric" leaves (fig. 1). 
There was no extension of vascular bundle sheaths into 
the mesophyll. Needle lifespans of cembran pine and 
Norway spruce were 9 and 12 years for the north-facing 
slope and 8 and 9 years for the east-facing slope, respec- 
tively (fig. 2). The proportion of older needles of both spe- 
cies was larger on the north-facing slope than on the east- 
facing slope. In Norway spruce, the frequency of younger 
needles on the north-facing slope was smaller than that 
on the east-facing slope. A negative correlation between 
the Ught-saturated photosynthesis and needle longevity 
was found in both species (fig. 3). 



Physiological Parameters 

There was a positive correlation between the stomatal 
conductance and the light-saturated photosjoithesis in 
both species (fig. 4). No difference between the north- and 
east-facing slopes was foimd. The quantum yield and car- 
boxylation efficiency (CE) of both species on the north- 
facing slope tended to be lower than of those on the east- 
facing slope (table 1). 

The maximimi photo sjTithe sis at light and CO^ satura- 
tion (Pj^^) of both species on the north-facing slope was 
lower than that on the east-facing slope (P < 0.05). The 
CE of cembran pine was sHghtly higher than that of 
Norway spruce. The nitrogen concentration in needles 
of cembran pine was higher than in needles of Norway 
spruce. However, there was no statistical difference in 
both species between the slopes. 

DISCUSSION 

The estimation of is based on the uniform response 
of the stomata of a leaf (Terashima 1992). The anatomy 
of needles of cembran pine and Norway spruce was simi- 
lar to the homobaric leaf. Non-uniform stomatal aperture 
in needles of loblolly pine was observed only when needles 




Figure 2 — Frequency of needle age in percent of Figure 3 — Photosynthesis in relation to needle age 

total needles of cembran pines (right) and Norway in cembran pines (left) and Norway spruces (right) 

spruces (left) growing on east-facing and north- growing on east-facing and north-facing slopes at 

facing slopes at timberline. timberline. 



79 



8 



sz 

c 

w 
o 

o 
sz 



'E 

-54 

E 



1 


1 

• North-facing slope 


o 


uusi-Taciny siop6 






1 


1 



0 40 80 0 40 80 

Stomotal conductance (mmolrfi^s^) 

Figure 4 — Photosynthesis in relation to stomatal 
conductance in cembran pines (left) and Norway 
spruces (right) growing on east-facing and north- 
facing slopes at timberline. 



were treated with abscisic acid, not with the low tempera- 
ture in roots (Day and others 1991). According to these 

facts, we can estimate the CE calculated with C. values. 

1 

The P^^ from needles of the north-facing slope was 
smaller than of those of the east-facing slope, while needle 
lifespan of the north-facing slope increased. Many physi- 
ological parameters in needles from the north-facing slope 
were almost the same as those on the east-facing slope. 
However, slightly lower P^^ in both species from the 
north-facing slope may be attributed to the lower soil 
temperature (Day and others 1991; Hasler 1982, Turner 
and others 1982). The lower P of needles from the 

max 

north-facing slope may be compensated for by the longer 
lifespan of needles. Tree growth on the north-facing slope 
may be retarded by the shortage of radiation and low soil 
temperatures. 

ACKNOWLEDGMENTS 

We thank Dr. H. Turner for providing the climate 
data of the Stillberg site. Prof. W. K. Smith for collecting 
references, and Mrs. M. J. Sieber for editing the English 
version of this manuscript. 



REFERENCES 

Davies, C. E.; Benecke, U. 1980. Fluidized bed coating 
of conifer needles with glass beads for determination 
of leaf surface area. Forest Science. 26(1): 29-32. 

Day, T. A.; Heckathorn, S. A.; DeLucia, E. H. 1991. Limi- 
tations of photosynthesis in Pinus taeda L. (loblolly 
pine) at low soil temperatures. Plant Physiology. 96: 
1246-1254. 

Hasler, R. 1982. Net photosynthesis and transpiration 
of Pinus montana on east and north facing slopes at 
alpine timberline. Oecologia. 54: 14-22. 

Koike, T.; Hasler, R.; Matyssek, R.; Item, H. Seasonal 
changes in the photosynthetic capacity of Larix decidua 
and Pinus cembra planted on contrasting slopes at the 
timberline at Stillberg (Davos), eastern Switzerland. 
Trees, submitted. 

Kuoch, R.; Amiet, R. 1970. Die Verjiingung im Bereich der 
oberen Waldgrenze der Alpen. Eidgenossische Anstalt 
fiir das forstliche Versuchswesen, Mitteilungen. 46: 
159-328. 

Livingston, N. J.; Black, T. A. 1987. Stomatal characteris- 
tics and transpiration of three species of conifer seed- 
lings planted on a high elevation south-facing clear-cut. 
Canadian Journal of Forest Research. 17: 1273-1282. 

Schonenberger, W.; Frey, W. 1988. Untersuchungen ziar 
Okologie imd Technik der Hochlagenaufiforstimg. 
Forschungsergebnisse aus dem Lawinenanrissgebiet 
Stillberg. Schweizerische Zeitschrift fur Forstwesen. 
139: 735-820. 

Terashima, L 1992. Anatomy of non-uniform leaf photo- 
synthesis. Photos)aithesis Research. 31(3): 195-212. 

Turner, H.; Hasler, R.; Schonenberger, W. 1982. Contrast- 
ing microenvironments and their effects on carbon up- 
take and allocation by yoimg conifers near alpine tree- 
line in Switzerland. In: Waring, R. H.: Carbon uptake 
and allocation in subalpine ecosystems as a key to man- 
agement: Proceedings of the lUFRO Workshop, 
Pl.07-00, Corvalhs, OR: Oregon State University: 
22-30. 



Table 1 — Physiological parameters of cembran pine and Norway spruce growing on an east- and a north-facing slope of an 
avalanche gully at Stillberg (2,185 m a.s.l.) 



Parameter 



Cembran pine 



E slope 



N slope 



Norway spruce 



E slope 



N slope 



Quantum yield 0.014 ±0.004 
(mol COj'mol/quanta) 

Carboxylation efficiency 0.024 ± 0.009 
(^mol•m-2•^ba^^' 



(mol»m-2»s-^) 



8.30 ± 1.51 



Nitrogen concentration 
(mg»g-^) 12.26 ± 0.90 



0.009 ± 0.002 

0.022 ± 0.009 

5.38 ± 0.77 

13.27 ± 0.91 



ns 



ns 



(1) 



ns 



0.012 ± 0.006 0.006 ± 0.002 



0.017 ± 0.002 0.018 ± 0.008 



6.87 ± 1.61 



10.43 ± 1.87 



4.93 ±0.12 



9.26 ± 1.22 



ns 



ns 



(1) 



ns 



'P<0.05. 



80 



THE BROAD-LEAVED KOREAN PINE 
FOREST IN CHINA 

Luo Ju Chun 



Abstract — Korean pine (Pinus koraiensis) is a tree species of 
high economic value. Its natural distribution area, the character- 
istics of distribution in the broad-leaved Korean pine forest re- 
gion of China, the biological and ecological characteristics of Ko- 
rean pine, the forest types of Korean pine, and the succession 
pattern of the Korean pine forest communities have been studied. 
The results are significant to conservation and sustaining use of 
Korean pine. 



Korean pine (Pinus koraiensis) is well known for its 
high-quality timber that can be used for many different 
purposes. Undoubtedly, Korean pine is one of the most 
important tree species in Chinese forests. 

Except for a few pure Korean pine forests, the species 
often occurs in the "conifer-broad-leaved tree mixed-forest 
communities" vnth many species of deciduous broad- 
leaved trees and other conifers, which is the most repre- 
sentative vegetation formation in the northeastern humid 
zone in China. 

NATURAL DISTRIBUTION AREA 

Korean pine — a relict species from the Tertiary — 
mainly occurs in northeast Asia and in a narrow zone 
along the west coast of the Japan Sea. This zone includes 
the southern part of eastern Russia, the eastern part of 
northeast China, North Korea, and the center of Honshu 
in Japan. The main natural range of Korean pine, how- 
ever, is in China. Korean pine is a representative species 
in Dahurian flora, and also is the main forest tree species 
in the eastern momitains of northeast China. 

In China, the broad-leaved Korean pine forest is found 
from 40° 15' N. to 50°20' N. and from 126° E. to 135° E., 
east of Song-nen plain and north of Song-liao plain. In 
the south it reaches Dan dong, in the north Hei-Ho 
coimty. The whole sirea takes a crescent shape covering 
Lesser Xingan Ling, Wan Da Mountain, Zhang Guang Cai 
Ling, Lao Ye Ling, and Chang Bai Mountain. Most of 
these mountains are trending from northeast to south- 
west. They are characterized by gentle topography and 
low elevation, not higher than 1,300 m. There are many 
different tree species, including many quite valuable 
broad-leaved trees. Their exploitation began more than 
100 years ago. Their timber output still makes up one- 
fourth of total timber production in China. 

The forest region, which is influenced by the Japan Sea, 
is characterized by temperate-zone monsoon climate. The 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Luo Ju Chun is Professor of Ecology and Forest Management, Forestry 
Resource College, Forestry University, Beijing, China. 



warm and moist climate is favorable to growth of Korean 
pine. Annual average temperature is 0-6 °C; the growing 
season is about 4-5 months. Temperatvu"e svun (>10 °C) 
is about 2,500-3,200 °C. At higher latitude, temperature 
is low, vyinter is long, and simimer short. Winter covers 
5 months. January is the coldest month; the absolute 
minimum temperature is -40 to -30 °C. In July, which is 
the warmest month, mean temperature reaches 20-24 °C. 
Winter is characterized by thick snow cover and mean 
temperatures ranging from -28 to -14 °C. Annual pre- 
cipitation amoimts to 600-1,100 mm £ind decreases from 
south to north; maximum is in summer (June to August). 
The annual relative hiunidity is about 60-70 percent. 
Hiunidity and temperature are favorable to growth of 
Korean pine and other trees. 

The soil in Korean pine forests is dark brown forest soil 
vfith a thick hvunus layer, and it is very fertile. 

CHARACTERISTICS OF 
VEGETATION DISTRIBUTION 

Horizontal Distribution 

Because of the wide range of Korean pine forest, the cli- 
mate is quite different from north to south in the distribu- 
tion area. Consequently, some differences in flora and 
plant commvuiities become apparent. The distribution 
area can be divided into two subzones: northern and 
southern temperate. The demarcation is from Don-g Nin 
in the eastern part of Jing Bo Lake to Ji Lin district until 
its west boiindary; it seems to nm along the parallel of 
44 °N. The differences in natiiral conditions and vegeta- 
tion can be described as follows: 

1. In the northern subzone, climate is cold, the growing 
season is only about 100-120 days, and annual rainfall 
£imounts to 500-700 mm; the southern subzone is charac- 
terized by a warm and hvunid climate and a growing sea- 
son of 130-150 days. Annual precipitation is about 700- 
1,100 mm. 

2. In the northern subzone, more conifers typical of the 
cold-temperate zone (for example, Picea koraiensis, P. 
jezoensis, and Abies nephrolepis) occur in Korean pine 
forests. Pinus pumila can be found in the subalpine. In 
addition, Quercus mongolica is more common on expo- 
siu-es to Bun. Larix gmelinii forests, mixed with Betula 
platyphylla, occur in the lowlands. Vaccinium vitis- 
idaea, Ledum palustre, Vaccinium uliginosum, Betula 
fructicosa — representatives of Dahurican flora — are typi- 
cal of larch forest undergrowth. In the southern subzone, 
there are more representatives of Chang Bai flora, such as 
Picea jezoensis var. komarovii, Abies holophylla, and 
Larix olgensis. 



81 



3. In the southern subzone there are more plant species 
than in the north. The forest often consists of 30-40 woody 
species forming three layers. In the southern subzone 
there are twice as many broad-leaved tree species in 
Korean pine forests than in the northern subzone; the 
genus Acer, for instance, is represented by only five spe- 
cies in the north, compared to 12 species in the south. 
The percentage of valuable broad-leaved trees (such as 
Franus mandshurica, Juglans mandshurica, Phello- 
dendron amurense, Tilia amurensis, Ulmus propingua) 
in Korean pine forests is higher in the southern than in 
the northern subzone. These broad-leaved trees usually 
grow together with Betula costata and Acer mono and rep- 
resent more than 50 percent of the species of southern 
Korean pine forests. Moreover, Carpinus cordata and 
Fraxinus rhynchophylla are very commom in the south, 
but occur rarely in the north. In addition, there are more 
than 10 vine plant species in the southern subzone, but 
only five or six in the north. 



1400 -r 




0-^ 



Figure 1 — Vertical distribution pattern of veg- 
etation in Lesser Xin An Ling (southern expo- 
sures): 1a, River head-valley-conifer forest with 
larch, spruce, fir; 1b, Bank depression, broad- 
leaved forest with Ulmus propinqua, Fraxinus 
mandshurica, Pliellodendron amurense; 2, Oak 
forest {Quercus mongolica) mixed with Betula 
dahurica, Tilia mandshurica; 3, Broad-leaved 
Korean pine forest mixed with spruce, fir, 
Betula costata, Tilia amurensis, Ulmus 
laciniata; 4, Spruce-fir forest belt; 5, Subalpine 
Betula ermanii-Pinus pumila elfin forest belt. 



Vertical Distribution 

The vertical distribution pattern of vegetation in Lesser 
Xin An Ling (fig. 1) is a typical example of the northern 
subzone: 

1. River valley forest: It is composed of Picea jezoensis, 
Abies nephrolepis, Larix gmelinii, Juglans mandshurica, 
Ulmus propinqua, Populus ussuriensis, and other species. 

2. Oak {Quercus mongolica) broad-leaved forest belt 
(150-300 m): It is secondary forest that established after 
destruction of virgin conifer-broad-leaved mixed forest, 
and it occurs usually on shallow soils on s\in-exposed 
slopes. 

3. Broad-leaved Korean pine forest belt (250-650 m): 
Korean pine is the dominant species; spruces and firs oc- 
casionally occur; the broad-leaved trees make up less than 
30 percent of the species. The volume of standing timber 
is the highest (300-600 m^/ha) in the northern forest re- 
gion. Korean pine grows on exposures to south and south- 
west, on mountain ridges, and on hill tops. 

4. Spruce-fir forest belt (650-1,000 m): Picea korai- 
ensis, Picea jezoensis, and Abies nephrolepis are the main 
tree species. 

5. Subalpine moss-elfin forest belt (1,000-1,080 m): 
The climate is cold and humid; wind is strong. The elfin 
forest is composed of Betula ermanii. This species exhib- 
its dwarf growth forms not taller than 4-6 m. Pinus 
pumila grows in the forest understory. The forest floor 
is covered mainly by moss. 

The vertical distribution pattern of vegetation in the 
southern subzone can be illustrated by the situation of 
Chang Bai Mountain (fig. 2): 

1. Broad-leaved forest belt (250-500 m a.s.l.): It is 
formed by Quercus mongolica, Betula platyphylla, Betula 
dahurica, and Populus davidiana. It is a typical secon- 
dary forest following destruction of broad-leaved Korean 
pine forest. Locally, basswood, maple, elm, walnut, and 
corktree form mixed forests. At lower elevations, these 
broad-leaved forests are scattered. 

2. Broad-leaved Korean pine forest belt (500-1,200 m): 
Korean pine is the main species, and is mixed with vari- 
ous broad-leaved trees: Betula costata, Carpinus costata, 
Ulmus laciniata, Acer mono, Tilia amurensis, Quercus 
mongolica, especially the valuable broad-leaved tree spe- 
cies Ulmus propinqua, Phellodendron amurense, Juglans 
mandshurica, Fraxinus mandshurica, and, at valley bot- 
toms, some spruce and fir can be found. 

3. Spruce-fir forest belt (1,200-1,800 m): It is formed 
by Picea jezoensis, Picea koraiensis, and Abies nephrolepis, 
with additional Betula ermanii and Larix olgensis in the 
upper part of the belt, or with Korean pine, Betula 
costata, Ulmus laciniata, and Tilia amurensis in its lower 
part. 

4. Subalpine Betula ermanii mossy elfin forest belt and 
subalpine meadow belt (1,800-2,100 m): These belts are 
characterized by cold and windy climate, steep topogra- 
phy, and shallow soils. Betula ermanii forests alternate 
with vast meadow areas. 

5. Alpine tundra belt (>2,100 m): Vegetation consists 
of different dwarf shrub species, mosses, and lichens. 



82 




3000 -r 



2500 



2000 - 



(0 

5 1600 - 
0) 



1000 - 



500 - 



0-^ 



Figure 2 — Vertical zonal spectrum of vegetation 
in Chang Bai Mountain: 1 , Secondary deciduous 
broad-leaved forest; 2, Conifer-broad-leaved 
mixed forest (broad-leaved Korean pine forest); 
3, Coniferous forest belt: 3a, Spruce-fir-Korean 
pine forest subbelt; 3b. Spruce-fir coniferous for- 
est subbelt; 4, Subalpine Betula ermanii-e\i\n for- 
est belt, and subalpine meadow belt; 5, Alpine 
tundra belt. 



As far as we know, the higher the latitude the lower is 
the upper limit of Korean pine. In the southern subzone 
Korean pine occurs up to 1,200 m a.s.l.; in the north its 
uppermost altitudinal limit is located at about 650 m. 

BIOLOGICAL AND ECOLOGICAL 
CHARACTERISTICS 

Korean pine is a long-lived species. Its lifespan is 300- 
400 years, sometimes 500 years. Tree height usually 
reaches 30 m. The diameter can reach 1 m. Korean pine 
is one of the valuable sawtimber trees in the world. It 
grows slowly in the first 5-8 years. After the age of 10 its 
growth rate increases rapidly. Its diameter growth begins 
to accelerate at the age of 12. The height growth becomes 
more rapid after 16-18 years. Accelerated growth contin- 
ues vmtil the age of 30. Korean pine is a tree species 
growing fastest at middle age. In general, growth rate of 
Korean pine increases earlier in plantations than in natu- 
ral forests. In clear cuttings, the radial growth of 8-year- 
old trees is 2.6 times higher than in nattu-al forests, and 
height growth is twice as much as in trees at the same 
age that became naturally established. 



Secondary growth of Korean pine is common in autiunn, 
but it appears mainly in trees less than 10 years old. 
After 80 years growth, Korean pine reaches a height of 
18-20 m. At this age the tree top often begins to fork and 
bear cones. Obviously, the forking is closely related to 
cone production. 

Korean pine has shallow roots. Its taproot is imderde- 
veloped and degenerates easily. Thus, seed trees left on 
clear cuttings are sensitive to windbreak. 

Most Korean pine trees begin to produce cones at the 
age of 80-100 years. Nevertheless, cones can also be 
found on 30-year-old trees. In plantations, however, 
even 20-year-old trees are able to bear cones. 

There are two ecot5T3es of Korean pine: one is the 
leptodermis form displaying thin bark with small and 
shallow scale or long strip lobes. It forks less, grows 
faster in height, and produces better timber. The other 
is the pachidermis form characterized by deep and large, 
long, square-shape lobes on the tree trunk. It forks more, 
grows slower in height, and the timber quality is not as 
good as in the leptodermis type. The leptodermis type 
should be used for afforestation. 

Korean pine belongs to thermophile species; its vegeta- 
tive growth requires 6-7 °C of the lowest average tempera- 
tiu-e and the most suitable temperature is 14-16 °C in the 
yoimg stage. The relative moistvu^e it demands is 70 per- 
cent and, during the period of most rapid growing, rela- 
tive hvunidity of about 70 percent is required. Korean 
pine is intolerant to high temperature during vegetative 
growth. If average temperature exceeds 15-16 °C, height 
growth will decrease or even stop. On the other hand, it 
will not be affected by winter temperatiires as low as 
-50 °C. 

At seedling stage a little shading is favorable to Korean 
pine, althovigh it can tolerate full siinlight. Its light re- 
quirement increases with age. While 3 years old, seed- 
lings require 60-70 percent of incoming solar radiation. 
After the fifth year, young trees become totally shade 
intolerant. 

Korean pine grows best on thick, moist, fertile, and well- 
drained soils with pH 5-6, but it can also exist on shallow 
dry soil, poor in nutrients. It grows badly on soils rich in 
clay. However, Korean pine requires appropriate soil mois- 
ture. In bogs and on sites with stagnant water, growth is 
hampered, and Korean pine is rarely seen there. 

PRINCIPAL FOREST TYPES 

Mu Dan Jiang forest region, the distributive center of 
Korean pine, may serve as an example to classify Korean 
pine forest into the following forest types: 

Steep-Slope Carex calltnichos-Korean Pine 
Forest — This type occurs on simny exposures or narrow 
ridges at an altitude of 800-1,000 m. These usually are 
poor sites. The conmiunity can be divided into two layers. 
The forest canopy is formed by Korean pine (90 percent) 
and Quercus mongolica or Tilia amurensis (10 percent), 
which have a higher crown density than Korean pine. In 
the second layer there is some Korean pine, spruce, fir, 
and linden. The growing stock is 400-500 m^/ha. The 
dominant species on the forest floor are Carex calltnichos 
and C. ussuriensis. 



83 



CoryluS'Carex-Korean Pine Forest— This type is 
found on mountains with gentle slopes at altitudes of 
700-900 m. As to favorability to tree growth, these sites 
can be considered intermediate. Korean pine prevails 
(60-70 percent) followed by Tilia amurensis (20 percent), 
and some Abies nephrolepis, Picea jezoensis, Ulmus 
propinqua, Acer mono, and Betula costata. The growing 
stock is about 400-500 m^/ha. There are more tree species 
associated, such as Acer sp. There is more undergrowth, 
and the main species is Corylus mandshurica. On the for- 
est floor about 30 species are to be found. Carex sidero- 
sticta and ferns are the most common plants. 

Fern-Spruce-Fir-Korean Pine Forest — This type 
occurs on the lower gentle slopes (inclination about 10°) 
of the mountains or along side streams at an altitude of 
500-800 m. The community construction is complex. The 
canopy layer is formed by Korean pine and some Populus 
ussuriensis and Picea jezoensis. The second layer is com- 
posed of Picea jezoensis and P. koraiensis (50 percent), 
Abies nephrolepis (20 percent), and Tilia amurensis 
(20 percent). Other species are Betula costata, Ulmus 
laciniata, Acer mono, Phellodendron amurense, Juglans 
mandshurica, Fraxinus mandshurica, and many kinds of 
maple. Because of better site conditions, growing stock is 
higher (250-350 m^/ha for the first story and 200 m^/ha for 
the second story). The undergrowth is abundant with 
over 20 species; about 30 species occur on the forest floor. 
Ferns are dominant (such as Dryopteris). 

Fern-Moss Korean Pine Forest — This type is found 
in the lower part of mountains or in broad valleys; slope 
inclination is 0-5°. Two stories can clearly be distin- 
guished: Pinus koraiensis (80 percent) and a few Picea 
jezoensis, Fraxinus mandshurica, Ulmus propinqua, Acer 
mono, and Tilia amurensis form the forest canopy. In the 
second story there is more Picea jezoensis, Betula costata, 
Acer mono, Abies nephrolepis, Tilia amurensis, and a few 
Juglans mandshurica, Fraxinus mandshurica, Ulmus 
propinqua, Betula platyphylla, and many kinds of maple. 



So the tree species are very abundant. This forest type 
is typical of the sites most favorable to tree growth. The 
growing stock can reach 600 m^/ha. The imdergrowth 
mainly consists of Lonicera sp., Viburnum sp., and Sor- 
baria, and is well developed. Carex and multiple fern and 
luxuriant moss cover are common throughout the area. 

COMMUNITY SUCCESSION 
PATTERNS 

Two succession patterns for the broad-leaved Korean 
pine communities are obvious: 

Succession Without Disturbance — Natural regen- 
eration is poor under virgin Korean pine forest cover, al- 
though a comparatively large quantity of seedlings may 
appear after a seed year. However, they will only be able 
to grow up if the canopy is cut thin. Because Korean pine 
lives longer than other coniferous and broad-leaved trees, 
and tolerates shade for longer period, it tends to form 
multigeneration and multistoried forests. 

Succession After Disturbance — ^After destruction of 
broad-leaved Korean pine forest the succession goes on as 
shown by the following sequences: 

Selective cutting -» conifer-broad-leaved mixed forest 
in which main elements are broad-leaved trees. Protec- 
tion broadleaved-Korean pine forest. 

Clear cutting or fire -> secondsiry bare land. Protec- 
tion grass and shrub communities conifer-hardwood 
mixed forests in which main elements are hardwood trees 
-> conifer-broad-leaved mixed forest in which main ele- 
ments are coniferous trees broad-leaved-Korean pine 
forest. 

To promote the reestablishment of Korean pine forest, 
management should be adjusted to this succession 
process. 



84 



CLIMATES WHERE STONE PINES 
GROW, A COMPARISON 

Tad Weaver 



Abstract— While stone pine climates are similar, species 
adapted to relatively moderate climates may be excluded from 
the ranges of congeners by more severe climates, and species 
with longer warm-moist growing seasons are probably more pro- 
ductive than congeners. Absolute low/summer average/absolute 
high temperatures for stone pines listed in order of increasing 
absolute low temperature are Pinus sibirica (-55/13/37 °C), 
P.pumila (-52/9/36 °C), P. koraiensis (-42/11/36 °C), P. albi- 
caulis (-34/9/29 °C), and P. cembra (-23/8/27 °C). The Walter 
drought index shows little stress in stone pine forests despite 
large differences in summer/winter precipitation: in order of 
increasing summer rainfall, precipitation is P. albicaulis (102/ 
829mm), P. pumiVa (142/264mm), P. sibirica {187 /245mm), 
P. cembra (323/616mm), P. koraiensis (394/242mm). Estimated 
thawed-soil growing season increases from P. albicaulis (4.5mo), 
through. P. pumila (4.6mo), P. sibirica (5.5mo), and P. cembra 
(6.3mo) to P. koraiensis (7.8mo); growing seasons of the first 
three trees could be shortened by drought. 



Stone pines grow in most of the boreal zone (fig.l) 
(Fullard and Darby 1964; Lanner 1990; Mirov 1967). 
Pinus pumila and P. sibirica occupy vast far-northern 
(50 to 70° N. latitude) areas in northeastern Asia; P. sibir- 
ica ranges from just west of the Urals (55° E. longitude) 
halfway to the Pacific (115° E. longitude) and P. pumila 
ranges from 115° E. longitude to the Pacific (165° E. longi- 
tude). P. koraiensis grows at the boreal-deciduous forest 
transition in eastern China (120 to 150° E. longitude and 
45 to 55° N. latitude with outliers to 35° N. latitude at 
high altitude). Paralleling P. koraiensis, P. albicaulis 
occupies the Rocky-Cascade-Sierra Mountain chains of 
North America from 55° N. latitude southward to 45° N. 
latitude with outhers to near 35° N. latitude. Pinus 
cemba appears in the Alps at approximately 45° N. lati- 
tude. Glaciated parts of the conifer forest zone— both in 
northern Europe and North America— lack stone pines 
(fig.l). 

Due to their common ancestry and common boreal 
forest habitat, one expects the trees to occupy similar cli- 
mates. This paper tests that hypothesis by comparing cli- 
mates occupied by the trees with respect to 20factors that 
may be important in determining the ranges and produc- 
ivities of the species. The information may be especially 
useful for predicting the success of introductions— from 
one region to another— of stone pines, their tree associates, 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Statxis of Our Knowledge, St.Moritz, 
Switzerland, September5-ll, 1992. 

Tad Weaver is Plant Ecologist, Biology Department, Montana State 
University, Bozeman, MT 59717. 



and to some degree, even their herb, crj^jtogam, and ani- 
mal associates. Such introductions might be of interest 
as either producers or pests. 

METHODS 

Climates of environments dominated by closely related 
stone pines (Lanner 1990; Mirov 1967) were compared. 
The comparisons are nonstatistical, since the objective 
was to discuss biological responses rather than the cli- 
mates themselves. 

The climates were characterized by choosing three to 
four stone pine stands at which temperature and precipi- 
tation were regularly measured (a complete sample of 
accessible data); summarizing the data for periods of 10 
years (P. albicaulis and P. cembra) or longer (unspecified); 
and calculating means for selected parameters. The sta- 
tions studied are Usted here— from west to east— each 
with its approximate latitude, longitude, and altitude (m). 
Note that P. albicaulis and P. cembra grow significantly 
higher than, and to the south of, the Asian pines. Aster- 
isked Siberian stations lack most temperature data. Sci- 
entists who helped identify weather stations in stone pine 
climates are listed in the acknowledgments. 

Pinus albicaulis stations (USDC 1961-80) were Crater 
Lake, OR (43° N., 122° W., 1,990m), Ellery Lake, CA 
(38° N., 119° W., 2,940m), Old Glory, BC (49 ° N., 119° W., 
1,008m), and Kings Hill, MT (46 ° N., 110° W., 2,225m). 
Pinus cembra stations (contributed by W.Tranquillini) 
were Haggen (47° N., 11° E., 1,800m), Obergurgl (47 ° N., 
11° E., 2,070m), and Patscherkofel (48 ° N., 11° E., 
1,952m). Pinus sibirica stations (Muller 1982) were 
Serov (60° N., 61° E., 132m*), Surgut (61 ° N., 73° E., 
40m), Kolpasevo (58 ° N., 83° E., 76m*), Jenisejsk 
(58° N., 92° E., 78m*), Krasnojarsk (56 ° N., 93° E., 
15m), Tura (64 ° N., 100° E., 130m*), Irkutsk (52 ° N., 
104° E., 468m), and Kirensk (58 ° N., 108° E., 256m). 
Pinus pumila stations (Muller 1982) were Vitujsk (64° N., 
122° E., 107m*), Jakutsk (62 ° N., 130° E., 100m*), 
Verchojansk (68° N., 133° E., 137 m), Ochotsk (59° N., 
143° E., 6 m), Zyranka (65° N., 151° E., 43m*), 
Petropavlovsk-Kamcatskij (53° N., 159° E., 32m), Apuka 
(65° N., 170° W., 10m*), and Anadyr (65 ° N., 178° E., 
62m). Pinus koraiensis stations (contributed by Luo 
Ju Chun and Zhao Shidong) were AnTu (43° N., 128° E., 
591m), Chargbei-ShenYang (42 ° N., 128° E., 738m) and 
YiChun (48° N., 129° E., 231m). 

Parameters studied describe climates with respect to 
temperature, precipitation, and growing season. The fol- 
lowing three paragraphs list these parameters and outline 
the rationales for their choice; the rationales are devel- 
oped further in Weaver (1993) and in the following 
discussion. 



85 



































B 




7^ — 


•1 





Figure 1 — Distribution of stone 
pines in relation to space, conifer 
forests, and pleistocene glaciation. 
North America and Eurasia are 
seen from a polar view. Shading 
locates the conifer zone on the two 
continents (Fullard and Darby 
1964). Areas within the dotted lines 
received pleistocene glaciation 
(Denton and Hughes 1981). 
Ranges of the stone pines (Mirov 
1967) are outlined by solid or 
dashed lines: A = Pinus cembra, 
B = P. sibirica, C = P. pumila, 
D = P. koraiensis, E = P. albicaulis. 
The absence of stone pines in the 
conifer zone of northwestern 
Eurasia and parts of North America 
could be due to glaciation. 



Ten temperature parameters were compared. Midwin- 
ter frost danger was described by the long-term minimum 
temperature (the absolute low), the average January 
minimum, and the average January maximum. Extreme 
fall and spring frosts are represented by the absolute low 
in the first and last months of winter (that is, the absolute 
lows in the months with OC average air temperature; 
Weaver 1994). The average growing-season temperature 
was calculated across those months when average air 
temperatures were above 0 °C. This average was recalcu- 
lated using "temperature growth support units = Q" to 
account for the rise in rates of metabolic processes with 
rising temperature; these points lie on a curve defined 
by 0 °C = 0, 1 °C = 1, 11 °C = 2, 21 °C = 3, 31 °C = 4, and 
41 °C = 5 (Weaver 1994). Summer highs were represented 
by the average July minimum, average July maximum, 
and long-term high. 

Seven precipitation parameters were compared. 
October- June precipitation was measured to determine 
whether soil was moist in winter and at the opening of 



the growing season. Summer precipitation was indexed 
by July-September precipitation and by precipitation in 
the wettest and driest month in that period. Since plants 
require water, not precipitation, months with a positive 
water balance were estimated with the Walter index 
(Nielson 1986; Walter 1973), which assumes that for ev- 
ery 2 °C rise in average temperature 1 mm of precipita- 
tion will be evapotranspired. Two derivatives were stud- 
ied: drought months (duration of periods with a negative 
water balance) and drought magnitude (total estimated 
deficit). While the Walter index tends to underestimate 
drought (Stephenson 1990), it may be adequate in these 
cool climates. Poikilohydric (Larcher 1975) organisms 
may be more dependent on a number of growing hours set 
by the number of rain days than on total precipitation; 
rain days per month were therefore recorded. 

Four growing-season parameters were compared. Length 
of the growing season was indexed first as the number of 
months when air temperature is above 0 °C, that is, the 
approximate number of months when the soil is thawed 



86 



(Weaver 1994). Second, drought months (none according 
to the Walter index) were subtracted from the warm- 
season index to create a better warm-moist season index. 
Despite contrary observations (Weaver 1994), growth is 
expected to be indexed by the integral of temperature over 
growing season. Thus two indices of growing season pro- 
ductivity (W eaver 1994) were calculated: [growing season 
X (average growing season temperature -5 °C) (Chang 
1968)] and, considering the Q effect, [growing season 
X average growing season Q]. 

WINTER TEMPERATURES 

Average winter (January) temperatures in northeastern 
Asia are the world's coldest and they warm southward 
(Fullard and Darby 1964). Normal daily lows in regions 
occupied by the trees are Pinus pumila (-30 °C), P. sibir- 
ica (-27 °C), P. koraiensis (-27 °C), P. albicaulis (-14 °C), 
and P. cembra (-8 °C), respectively (table 1). Daily highs 
in winter average 8 °C higher (table 1). 

Absolute lows experienced by Pinus sibirica (-55 °C) 
and P. pumila (-52 °C) rise to P. koraiensis (-42 °C), 
P. albicaulis (-34 °C), and P. cembra (-23 °C) (table 1). 
Pinus albicaulis and P. cembra may be protected from ex- 
treme low temperatures by the drainage of cold air from 
their mountaintop sites. Becwar and Burke 1982 show 
that timberline conifers of the Colorado Rockies do not 
survive temperatures below -40 °C; thus P. sibirica and 
P. pumila may be the only stone pines tolerant of -50 °C. 

Frosts of early fall and late spring may catch trees in 
partially hardened states. Absolute lows in the first and 
last months of winter were -21 °C for P. koraiensis, -17 °C 
for P. sibirica, and -10 °C for the remaining trees. Why 
are absolute lows for fall-winter-spring frosts lower in the 
P. koriaensis and P. sibirica regions than in the P. pumila 
regions where average temperatures are lower (table 1; 
Fullard and Darby 1964)? 



SUMMER TEMPERATURES 

Continentality allows northeastern Asia to warm in 
summer more than less continental parts of the polar 
region. Thus absolute maximum temperatures of Pinus 
sibirica (37 °C) and P. pumila (33 °C) are considerably 
higher than those of P. cembra (27 °C) and P. albicaulis 
(29 °C) (table 1). Absolute summer highs in the P. korai- 
ensis forest at the boreal-deciduous forest transition are 
as high (36 °C) as in interior forests. Average July highs 
in P. sibirica and P. koraiensis forests are like those in 
Rocky Mountain Douglas-fir ( Pseudotsuga menziesii ) for- 
ests, a vegetation zone below the Rocky Mountain stone 
pine zone (Weaver 1994). 

Daily growth rates are determined by average tempera- 
ture conditions in the growing season. (1) Temperature 
conditions are represented most simply by average tem- 
perature; this is 13 °C for Pinus sibirica, 11 °C for P. korai- 
ensis, and 8 to 9 °C for P. pumila , P. albicaulis, and P. cem- 
bra. Like July highs, 13 °C average temperatures, are 
similar to those found in the Douglas-fir/ponderosa pine 
forests of the Rocky Mountains (Weaver 1994). (2) Be- 
cause growth rises exponentially with increasing tem- 
perature (Qjq; Larcher 1975), a better index of tempera- 
ture on growth may be a cross-season average of growth 
support units (Weaver 1994), which give greater weight 
to high than low temperatures. With this index climates 
cool from P. sibirica to P. albicaulis to P. pumila-P. korai- 
ensis to P. cembra (table 1). 

PRECIPITATION 

Winter-spring precipitation is high in Pinus albicaulis 
(829 mm) and P. cembra (616 mm) and lower (about 
250 mm) in northeastern Asia (table 2). Since the lower 
amounts will saturate most moimtain soils (Weaver 1978), 
the excess is expected to run off, to have little effect on 
water supplies in forest stands during the growing season. 



Table 1 — Temperatures 


^ in stone pine communities of the world. 


Climates are listed in order of increasing winter temperatures 






Species, location, number of stations 


2 






P . pumila 


P. sibirica 


P. koraiensis 


P. albicaulis 


P. cembra 


Temperature 


N. China 


Siberia 


Korea-China 


N. America 


Euro-Alps 


data 


4+ 


4+ 


3 


4 


3 






Winter Temperature 






Jan. mean min 


-30.0 ± 8.0 


-27.0 ± 2.0 


-27.0 ± 2.0 


-14.0 ±2.0 


-8.0 ± 0.0 


Jan. mean max 


-24.0 ± 9.0 


-19.0 ±2.0 


-11.0 ±2.0 


-5.0 ± 3.0 


-1.0 ± 1.0 


Abs min 


-52.0 ± 4.0 


-55.0 ± 2.0 


-42.0 ± 1.0 


-34.0 ± 2.0 


-23.0 ± 1 .0 


Frost spring 


-11.0 ±3.0 


-18.0 ±2.0 


-26.4 ± 0.2 


-1 1 .0 ± 1 .0 


-10.0 ± 1.0 


Frost fall 


-11.0 ±3.0 


-16.0 ±2.0 


-27.6 ± 0.7 


-1 1 .0 ± 1 .0 


-10.0 ± 1.0 






Summer Temperature 






July mean min 


8.0 ± 0.0 


12.0 ± 1.0 


14.0 ±1.0 


4.0 ± 1.0 


5.0 ± 1.0 


July mean max 


15.0 ±2.0 


21.0± 1.0 


26.0 ± 0.0 


18.0 ± 1.0 


14.0 ±1.0 


Abs max 


33.0 ±1.0 


37.0 ± 1.0 


36.0 ±2.0 


29.0 ± 1.0 


27.0 ± 2.0 




9.0 ± 1.0 


13.0 ±1.0 


11.3± 1.2 


9.0 ±1.0 


8.0 ± 1.0 




1.9 ±0.2 


2.4 ± 0.1 


1.9 ±0.1 


2.1 ±0.2 


1.7 ±0.1 



'Temperature data (°C) are the mean ± one standard error. Absolute temperatures are recorded for 1 0 years in P. albicaulis and P. cembra; 
records for the Asian pines are unspecified (presumed longer). T^, and Q^, are growing season averages defined in the text. 

^ Sample size is four for P. albicaulis (except average max and min for January and July, n = 5), three for P. cembra, and for both P. pumila and 
P. sibirica eight, except for January max-min, July max-min, and spring-fall frost temperatures where n = 4. 



87 



and, thus, to have httle effect on production. Large snow- 
falls in the P. albicaulis and P. cembra forests, relative 
to those in the Asian forests, surely result in greater 
snowpacks, which shelter ground-level plants and ani- 
mals less frost tolerant than the trees. 

Summer precipitation is lower where moisture carrying 
air masses cross mountains (northeastern Asia and the 
Rocky Mountains) than where they do not (Alps and east- 
ern China). Thus summer rainfall increases from Pinus 
albicaulis (102 mm) to P. pumila (187 mm), P. sibirica 
(187 mm), P. cembra (323 mm), and P. koraiensis 
(394 mm). Precipitation in the driest month parallels 
summer rainfall (table 2). 

Soil water may be adequate in spite of low rainfall if 
evapotranspiration is low. If we assume that 1 °C degree 
evaporates 2 mm per month (Nielson 1986, 1992; Walter 
1973), we see that stone pines do not experience drought 
in any month (table 2). In addition, some buffering 
against drought occurs because snowmelt water stored in 
the soil provides a supplement to summer showers. How- 
ever, even at boreal temperatures Walter's index may un- 
derstate evaporation (Stephenson 1990), so there is prob- 
ably less summertime restriction of growth by drought 
(stomate closure) in Pinus cembra and P. koraiensis than 
in P. albicaulis, P. sibirica, and P. pumila forests. 

Organisms without water reserves — such as lichens, 
mosses, and invertebrates — are more sensitive to summer 
drought and may distinguish wetter and drier forests. 
These organisms are surely inhibited in Pinus pumila 
and P. albicaulis forests (one to eight rain days/month) 
relative to P. cembra and P. koraiensis forests (16 to 18 
rain days). 

GROWING SEASON 

Survival depends on photos3aithetic provisioning for 
winter respiration and the outcome of competition may 
depend on excesses above this basic provisioning. One 
expects production to be correlated with the number of 
warm-moist days, with the warmth of those days ( Q^^, 
Larcher 1975), and thus with their product. 



If water stress never occurs, the length of the warm- 
moist season increases from Pinus albicaulis (4.5 mo) 
to P. pumila (4.6 mo), P. sibirica (5.5 mo), P. cembra 
(6.3 mo), and P. koraiensis (7.8 mo) (table 3). And if tem- 
perature conditions were identical in these forests, one 
would therefore expect production in P. cembra and 
P. koraiensis forests to be half again as great as in 
P. pumila and P. albicaulis forests. Actual differences 
could be even larger because the temperature-defined 
growing season may overstate stand productivity in drier 
regions. For example, while subsoils remain moist all 
summer in higher (J. Brown, personal commimication) 
and lower (Weaver 1974) parts of the P. albicaulis zone, 
drying of surface soils apparently causes tree water stress 
at lower (B. Keane, personal communication), but not 
higher (J. Brown, personal communication), sites. Similar 
droughty periods probably occur in warmer P. pumila 
and P. sibirica forests. 

Production is expected to rise exponentially with rising 
temperature, and thus productivity might be better corre- 
lated with the product of season length and temperature 
"growth supporting units." While this production hy- 
pothesis was rejected in a cross-vegetation zone analysis 
(Weaver 1994), the genetic similarity of stone pines might 
allow it to operate here. If so, and if water deficits in 
Siberia do not shorten the growing season, P. koraiensis 
and P. sibirica will be promoted to the most productive 
stone pines (table 3). 

CONCLUSIONS 

From a tree's point of view, average conditions in the 
stone pine zones are similar: water stress is slight or 
nonexistent and growing season temperatures average 
9 to 13 °C. Seasonal extreme conditions may, however, 
prevent reciprocal transplantation. Winter or fall-spring 
frosts in the Pinus sibirica/P. pumila/P. koraiensis region 
may exclude the other pines. And weak droughts in the 
P. albicaulis/P. pumila/P. sibirica regions might exclude 
P. cembra or P. koraiensis. 



Table 2 — Precipitation^ in stone pine forests of ttie world. Climates are listed in order of increasing summer precipitation 







Species, location, number of stations ^ 






P . albicaulis 


P. pumila 


P. s ib irica 


P. cembra 


P. koraiensis 


Precipitation 


N. Annerica 


N. China 


Siberia 


Euro-Alps 


Korea-China 


data 


4 


4+ 


4+ 


3 


3 


Total 


931 ± 229 


407 ±137 


432 ± 21 


939 ± 9 


636 ± 68 


Oct.-June 


829 ± 234 


264 ± 76 


245 ± 22 


616 ±43 


242 ± 62 


July-Sept. 


102 ± 14 


143 ± 43 


187 ±11 


323 ± 36 


394 ± 17 


Wettest summer month 


116± 16 


165 ± 25 


181 ± 8 


214± 15 


346 


Driest summer month 


4± 4 


4± 2 


8± 2 


45 ±16 


32 


Summer drought months 


0± 0 


0± 0 


0± 0 


0± 0 


0± 0 


Summer water deficit 


0± 0 


0± 0 


0± 0 


0± 0 


0± 0 


Summer rain days, 












average number 


8± 1 


1 ± 1 


14± 1 


16± 1 


18 



'Precipitation (mm) data are total (sum of all months), winter, summer (July, August, plus September), driest summer month (July-September) recorded, 
wettest summer month (July-September) recorded, and average number of showers in June-September. High variances in total precipitation for P. 
albicaulis and P. pumila are reduced to 705 ± 51 mm and 274 ± 37 mm by omission of the Crater Lake and Petropavlovsk stations, respectively. 

^Sample size is four for P. albicaulis, three for P. cembra, three for P. Koraiensis, and eight for both P. pumila and P. sibirica, except for summer drought 
months and deficit, where it is only four. 



88 



Table 3 — Stone pine productivity correlates: growing season ,^ growing season temperature,^ and their products. Climates are listed in 
descending order by one estimate of productivity, warm moist season 

Species, location, number of stations 

P. koralensis P . cembra P. siblrica P. pumlla P. albicaulis 

Korea-China Euro-Alps Siberia N. China N. America 









4 


4 


4 


Warm season 


7.8 ±0.1 


6.3 ± 0.2 


5.5 ± 0.2 


4.6 ± 0.4 


4.5 ± 0.3 


Summer drought months 


0.0 ±0.0 


0.0 ± 0.0 


0.0 ±0.0 


0.0 ±0.0 


0.0 ±0.0 


Warm-moist season 


7.8 ±0.1 


6.3 ± 0.2 


5.5 ± 0.2 


4.6 ± 0.5 


4.5 ± 0.3 


k 

GS X Tg -5 °C 
GSxQg, 


11.3±1.2 


8.0 ±1.0 


13.0 ± 1.0 


9.0 ± 1.0 


9.0 ± 1.0 


1.9 ±0.1 


1.7 ±0.1 


2.4 ±0.1 


1.9 ±0.2 


2.1 ±0.2 


56.5 


18.9 


44.0 


18.4 


18.0 


15.6 


10.7 


13.2 


8.7 


9.5 



'Warm-season and warm-moist season are two indices of growing season. Warm-season months occur after average monthly air temperature hses 
above 0 °C and before it falls below 0 °C. The warm-moist season Is the warm season minus any months in which T/2 is greater than P(mm), Walter 1973; 
"Walter drought" does not occur in stone pine regions. 

H'wo indices of growing-season temperature are provided. T^, is the average temperature in growing season months. T -5C is used as one index of 
growth support on the assumption that growth does not occur below 5 °C (Chang 1 968). Q is the average of temperatures weighted tor their growth- 
supporting capacity: 0 °C = 0, 1 °C = 1 , 1 1 °C = 2, 21 °C = 4 {Weaver 1993). 



Organisms filling other niches might find the climates 
dissimilar (Weaver 1990). Poikilohydric organisms and 
himaans would react differently to the relatively smnmer- 
rainy climates of the P. cembra and P. koraiensis regions 
than to the other stone pine climates. And organisms 
wintering imder snow — small mammals, insects, plants 
of small stature — are more protected from the extremes 
of winter cold in the P. cembra and P. albicaulis regions 
than in the other climates. 

ACKNOWLEDGMENTS 

I thank (1) W. Tranqmllini (University of Innsbruck, 
Austria) for providing data from his P. cembra weather 
stations, (2) J. Luo (Beijing Forestry University, China) 
and Z. Shidong (Inst. Appl. Ecol., Shenyang, China) for 
providing data from weather stations in P. koraiensis cli- 
mates, (3) H. Lieth (University of Osabuck, Germany), 
P. Gorchakovsky (Acad. Sci. Inst. Ecol., Ekaterinburg, 
Russia) and P. Khomentovsky (Kamchatka Inst. Ecol., 
Petropavlovsk, Russia) for help in selecting P. sibirica 
and P. pumila weather stations, and (4) nineteen North 
American ecologists already acknowledged (Weaver 1990) 
for help in selecting North American P. albicaulis stations. 
I am also grateful for support to the project provided 
through Intermountain Research Station contract #INT- 
92720-RJVA (W. Schmidt). 

REFERENCES 

Becwar, M.; Burke, M. 1982. Winter hardiness limitations 

and physiography of woody timberline flora. In: Li, P.; 

Sakai, A. Plant cold hardiness and freezing stress. New 

York: Academic Press: 307-324. 
Chang, J. 1968. Climate and agriculture. Chicago: Aldine. 

304 p. 

Fullard, H.; Darby, H. 1964. The imiversity atlas. London: 
George Philip and Son. 176 p. 

Lanner, R. 1990. Biology, taxonomy, evolution, and ge- 
ography of stone pines of the world. In: Schmidt, W.; 
McDonald, K., comps. Proceedings — symposium on 



whitebark pine ecosystems: ecology and management 
of a high-mountain resource. Gen. Tech. Rep. INT-270. 
Ogden, UT: U.S. Department of Agriculture, Forest Ser- 
vice, Intermountain Research Station: 14-24. 

Larcher, W. 1975. Physiological plant ecology. New York: 
Springer. 252 p. 

Mirov, N. 1967. The genus Pinus. New York: Ronald 
Press. 602 p. 

Muller, M. 1982. Selected climatic data for a global set 
of standard stations for vegetation science. The Hague, 
Netherlands: Junk. 306 p. 

Neilson, R. 1986. High resolution climatic analysis and 
southwest biogeography. Science. 232: 27-33. 

Nielsen, R. 1992. Toward a rule based biome model. Land- 
scape Ecology. 7: 27-43. 

Stephenson, N. 1990. Climatic control of vegetation distri- 
bution: the role of water balance. American Naturalist. 
135: 649-670. 

Walter, H. 1973. Vegetation of the earth in relation to cli- 
mate and ecophysiolgical conditions. New York: 
Springer. 237 p. 

Weaver, T. 1974. Root distribution and soil water regimes 
in nine habitat types of the northern Rocky Mountains. 
In: Marshall, J., ed. The belowgroimd ecosystem. Sci. 
Ser. 26. Fort ColHns, CO: Colorado State University, 
Range Science Department. 351 p. 

Weaver, T. 1979. Changes in soils along a vegetational 
(altitudinal) gradient of the northern Rocky Mountains. 
In: Youngberg, C, ed. Proceedings of the Fifth North 
American forest soils conference. Madison, WI: Soil Sci- 
ence Society America. 14-29. 

Weaver, T. 1990. Climates of subalpine pine woodlands. 
In: Schmidt, W.; McDonald, K., comps. Proceedings — 
s)rmposium on whitebark pine ecosystems: ecology and 
management of a high-moimtain resource. Gen. Tech. 
Rep. INT-270. Ogden, UT: U.S. Department of Agricul- 
ture, Forest Service, Intermountain Research Station: 
72-79. 

Weaver, T. 1994. Vegetation distribution and production 
in Rocky Mountain climates — wdth emphasis on 
whitebark pine [these proceedings]. 



89 



Growth Characteristics 




International Workshop 
St. Moritz 1 992 



90 



GROWTH OF SWISS STONE PINES 
THAT ORIGINATED FROM AND WERE 
PLANTED AT SEVERAL ALTITUDES IN 
THE AUSTRIAN ALPS 

Kurt Holzer 



Abstract — For the purpose of gene conservation, a plus tree 
collection of Swiss stone pine {Pinus cembra) was conducted. 
Scions of 185 trees and 82 open-pollinated seed samples were 
collected within the natural range of Swiss stone pine in Austria, 
In 1987, clonal height was measured at an age between 25 and 
30 years. Open-pollinated families were planted at three differ- 
ent sites. Height was measured after 18 years. Growth of graft- 
ings and open-pollinated families helped identify the altitude of 
their origin. When planted below timberline, growth decreased 
with increasing altitude. 



In 1957, the Department of Forest Tree Breeding and 
Genetics, Federal Forestry Research Institute, Austria, 
commenced a selection program of Swiss stone pine (Pi- 
nus cembra). The objectives of this program were to 
preserve genetic resovirces of superior trees in a clonal 
orchard and to assess the growth capacity of the clones 
by field performance of open-pollinated families. 

Within the natural range of stone pine, timberline di- 
vides the distribution of stone pine, due to environmental 
conditions, into two silvicultural zones: (1) production for- 
ests and (2) "Kampfzone." Hence, those aspects had to be 
considered for the program (Holzer 1963, 1976). 

MATERIAL AND METHODS 

Covering the majority of natural stands in Austria, 
185 superior trees were selected between 1958 and 1965 
(Holzer 1961, 1969). Within production forests scions of 
124 and above timberline ("Kampfzone") scions of 61 indi- 
viduals were collected. The altitudinal range was approxi- 
mately 250 m below and 230 m above timberline (be- 
tween 1,650 m and 2,200 m above sea level). In addition, 
open-pollinated seeds were harvested from 82 trees. 

Graftings were planted in a conservation orchard in 
Purkersdorf-Stadlhutte near Vienna at 400 m above sea 
level. Progenies were raised in the nursery of the insti- 
tute (Mariabrvmn Vienna, 220 m above sea level) for 
6 to 8 years and then planted at three different sites: 
(1) Purkersdorf-Stadlhutte; (2) GroBe Zirbenwiese (1,650 m 
above sea level); Seetaler Alpen, a swampy meadow 
within the natiiral range of stone pine, approximately 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Kurt Holzer is Professor of Forest Genetics, Institut fur Waldbau, 
Universitat fur Bodenkultur, A-1190 Vienna, Austria. 



250 m below timberline; and (3) Melcheben, Packalpe 
(1,700 m above sea level), close to timberline of Norway 
spruce (Picea abies). At sites 2 and 3, only 32 families 
were planted because of limited nvmabers of plants. 

In 1987, height was measvired at a clonal age between 
25 and 30 years. Heights of progenies were assessed at 
an age of 8 and 18 years, respectively. Exclusively de- 
scriptive statistics were employed. 

Some results have been previously published with 
respect to graftings (Holzer 1989) and progenies (Holzer 
1978). Recently, growth of vegetative and generative 
plant material has been measured (Feuersinger 1992). 

Since altitude of timberline is irregular within the 
Alps, comparisons are based on relative distances to local 
timberline. 

RESULTS AND DISCUSSION 

Results are summarized in figures 1-3 within the main 
distribution (about 150 m below to 100 m above timber- 
line); mean height of 30-year-old graftings decreased with 
increasing elevation of origin fi-om 3.8 m at 150 m below 
to 2.7 m at 100 m above timberline. Variation was high 
and yielded to 50 to 70 percent of the clonal mean. In 
general, the majority of graftings originating below tim- 
berline had superior growth. Presimiably, inherent 
growth capacity shows a cline within the main distribu- 
tion zone of stone pine (fig. 1). 

This growth pattern was also found in generative plant 
material. Progenies originating fi'om 100 m below timber- 
line showed pronoimced reduced growth. However, when 
growth was only evaluated within a single seed zone, it 
was always higher below the timberline than above (fig. 2). 
In the nursery, height of families (n = 50) originating fi'om 
production forests averaged 46.5 cm. Families in = 16) fi'om 
the "Kampfzone" had a slightly reduced growth of 40.0 cm 
at an age of 8 years. 

Transplants of families 18 years old showed different 
results. Clinal growth pattern is still distinguishable at 
site 1 (400 m above sea level) and site 2 (250 m below tim- 
berline). However, at site 3 (timberline) height growth re- 
actions were different. Here, families with small height 
growth potential at 400 m above sea level were superior. 
Figure 3 shows growth of families originating fi'om two 
different seed zones. In seed zone "Defi-eggen" growth 
of 18 families, 400 m and 1,700 m above sea level, was 
negatively correlated (r = -0.752***). Fovuteen families 
originating from seed zone "Prankerhohe" showed a con- 
sistent pattern. However, correlation (r = -0.205) was not 
significant. 



91 



200 -X 

150- 

100- 

? 

£ 50- 

O) 

O above 

f timber line 
below 

50 - 
100- 

150- 
200- 
250- 



o 

3 



O X X J X X S 

I 



fix X 



-|^T< 5-5-xW^X-X-^-> 



XXX X.J 

X X Sx X S X X X >j6^ I i ^ 

N 

XX Jgx X^xg)^ X 

I \ 

X xx8xi>x'*8 X X 

I X 

/ 

X X I XX X 

X S I X P XX X 

/ 

l/ 

I I I 
3 4 

Height of Graftings (m) 



Figure 1 — Total height growth of 30-year-olcl 
graftings In the plantation at 400 m above sea 
level with the altitude of their origin (distance 
from the local timberline). Dotted curve denotes 
the mean values. 



130 



120 



110 - 



100 



90 - 



1 ,700 m asl 

X 



o o 



X \ 

x"^ X 

X \ 

\ 



X 

o 



\ 



400 m asl 



100 



120 140 

Height (cm) 



160 



Figure 3 — The comparison of the mean of the 
1 8-year-old families on two different planting 
sites: 400 m and 1 ,700 m above sea level, re- 
spectively (provenance 1/6/100, Prankerhohe, 
marked with o, provenance 1/8/100, Defreggen, 
marked with x; see text also). Curve denotes 
correlation (r = -0.752) of the Defreggen zone. 



This observation is supported by a sowing trial 200 m 
above local timberline (2,100 m above sea level). Pro- 
genies originating from sites above timberline performed 
well; progenies from trees in production forests performed 
poorly (Holzer 1975). 



What conclusions regarding growth of stone pine might 
be drawn from these experiments so far? 

1. Height growth of stone pine shows clinal variation. 

2. When plants are grown in production forests, growth 
is positively correlated to elevation. 

3. Above timberline this correlation is negative. 



200- 
150- 
100- 



•^50- e 
c 

S above 
timber line 
o below 



•E 50- 
< 

100- 



4Q 0 

+ 



-X-XQW-^ 



1s 



o + ++ oma _pco>ooo 0 o o o 



+ + + Ot + 



150- 



200- 



■H-X |X X"xp X XX 



30 



-H-O-l- +■ 

I 

40 



50 60 

Height of Progenies (cm) 



70 



Figure 2 — Total height of the 8-year-old prog- 
enies (mean of families) of the harvested 
clones with the altitude of their origin (distance 
from the local timberline). Dotted curve de- 
notes the mean values. 



REFERENCES 

Feuersinger, P. 1992. Das Wachstum von Zirbe in 
Abhangigkeit vom Ursprungsort. Diplomarbeit, Univ. f. 
Bodenkultur, Wien. 140 p. 

Holzer, K. 1961. Der derzeitige Stand der Auswahl von 
Schwarzerlen und von Zirbe im Rahmen der 
Forstpflanzenziichtung. Informationsdienst 50, Allg. 
Forstztg. 72. 

Holzer, K. 1963. Pinus cembra L. as a pioneer at timber- 
line in the European Alps. Proc. World. Cons. Forest 
Genet, and Tree Imprv., Vol. 1 (FAO/FORGEN 1963, 
3/13, FAO, Rome). 

Holzer, K. 1969. Erste Ergebnisse von Zirbenauswahl- 
baumen {Pinus cembra L.). Centrbl.f.d. ges. Forstwes. 
86: 149-160. 

Holzer, K. 1975. Genetics of Pinus cembra L. Annales 

Forestales, Zagreb. 6(5): 139-158. 
Holzer, K. 1976. Breeding haploxylon pines for subalpine 

regions. XVI lUFRO-Congr. Oslo. Div. II: 216-227. 
Holzer, K. 1978. Versuchsanlage GroBe Zirbenwiese. Exk. 

Fiihrer Arbeitsgem. f. Hochlagenaufforstung. 5 p. 

[Mimeo.] 

Holzer, K. 1989. Drei Jahrzehnte Erfahrungen mit 
Zirbenpfropfimgen. Centrbl. f.d. ges. Forstwes. 106: 
79-88. 



92 



SEASONAL PATTERNS OF GROWTH 
AND PHOTOSYNTHETIC ACTIVITY OF 
PINUS PUMILA GROWING ON THE KISO 
MOUNTAIN RANGE, CENTRAL JAPAN 

Takuya Kajimoto 



Abstract — Seasonal patterns of growth and photosynthesis in 
Pinus pumila needles were investigated in relation to carbon 
uptake. Current needles elongated from July to late August. 
Photosynthetic activities of 2-year-old needles, both in sun- and 
shade-needles, reached seasonal maxima in August, while those 
of current needles did so in September. The results suggest that 
annual net carbon gain of P. pumila depended mainly on net pho- 
tosynthesis of old needles (1 to 4 year olds). 



Growth and production rates of trees within the timber- 
Une ecotone are generally restricted mainly due to the 
short growing season (Tranquillini 1979). However, most 
of the subalpine species show physiological adaptations to 
conditions associated with low temperature, lower photo- 
synthetic optimun temperature (Larcher 1975), and 
higher frost and freezing resistance (Sakai and Larcher 
1987). It has also been suggested that within krummholz 
mats consisting of Picea engelmannii and Abies lasiocarpa 
photosynthesis is enhanced by more favorable canopy- 
temperature conditions (Hadley and Smith 1987). Infor- 
mation on such physiological and structural functions is 
indispensable for understanding how alpine trees survive 
by achieving positive carbon balance during the short 
siunmers. 

In Japan, Pinus pumila Kegel, occurs in high-moimtain 
areas from central Honshu (the main island) to the north- 
em island of Hokkaido. This pine generally exhibits 
dwarfed growth forms, and regenerates by layering at 
the mature growth stage (Kajimoto 1992; Okitsu and Ito 
1984). The physiognomy of P. pumila is similar to Pinus 
mugo at the upper timberline in the European Alps or in 
the Carpathian Mountains (Holtmeier 1973, 1981; Wardle 
1977). 

The foliage biomass of P. pumila rsinges from 15 to 24 
tons d.w./ha despite the low tree heights. Pinus pumila 
stands develop dense canopies with leaf area density of 
about 5 mVm^, which is considerably more than in other 
conifer forests (Kajimoto 1989a). This canopy structure 
may result in large annual net production (Okitsu and Ito 
1989; Shidei 1963). However, seasonal changes in growth 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Takuya Kajimoto is a special Researcher of Science and Technology, 
Laboratory of Plant Production, Forestry and Forest Products Research 
Institute, Tsukuba, Ibaraki 305, Japan. 



and photosynthesis of P. pumila needles are still un- 
known in relation to their carbon uptake. 

In this paper, a process of photosynthetic production 
in P. pumila during the growing season is discussed 
based on observations of environmental factors and 
seasonal patterns of growth and photosynthesis in the 
needles. 

STUDY AREA 

The study area is located on the Kiso Mountain Range 
in central Japan. The range trends from south to north. 
The highest peak is Mount Kisokoma (2,965 m a.s.l.). 
In this mountain range, P. pumila trees naturally occur 
above 2,500 m. The study was carried out in the 
Shinshu University Experimental Forest (35°48' N., 
137°50' E.) in the northern part of the Kiso Mountain 
Range. The investigated pine stand is located on the 
northeast slope (15° inclination) at an altitude of 
2,600 m a.s.l. Dense canopy layer is developed between 
100 and 200 cm above the ground height (Kajimoto 
1989a). 

FIELD OBSERVATIONS 

Four siin-exposed terminal shoots of P. pumila were 
selected. Lengths of current shoots, current needles, 
and winter buds were measured seven times between 
late May and October 1988. Current needle length of 
each shoot was determined as the average of 10 fascicles 
selected at equidistant intervals along the shoot axis. 

At the plot, solar irradiance (400-1,100 nm in wave- 
length) above the canopy siuface was measiu-ed with a 
pyranometer sensor (LICOR, LI-200SB) at 1-hour inter- 
vals from June to October 1988. Photos3nithetic photon 
fliix density (PFD) (400-700 nm) above the canopy sur- 
face and air temperature inside the stand were also 
recorded at 1-hour intervals between mid-July and 
October 1988 using a quantum sensor (Koito, IKS-25) 
and a resistance thermometer (Koito, OPT-150). 

MEASUREMENT OF 
PHOTOSYNTHESIS 

Rates of COg exchange were determined on detached 
shoots that were collected 50 hours prior to the meas- 
urement. The shoots were carefully transported to the 
laboratory to minimize error associated with excision 
and transportation (Kajimoto 1990). An open system 



93 



with an infrared gas analyzer (Hartmann and Braun, 
URA-2) was used. The air was fed into an assimilation 
chamber (9 by 25 by 1 cm^) at a rate of O.Sl/min. COg con- 
centration of the input air was within the range of 360- 
400 ppm. The chamber was placed in a thermoregulated 
waterbath and illimiinated with incandescent lamps. 
PFD at the chamber surface was measured with a quan- 
tum sensor (LICOR, LI-190SB). Details of the system 
were described by Kajimoto (1990). 

Two terminal shoots (about 10 cm in length) of P. 
pumila were sampled from the sun-exposed canopy sur- 
face part (1.8-2.0 m) and in the lower shaded canopy lay- 
er (1.0-1.2 m) seven times between late May and October 
1988. At each sampling time, about 26 fascicles of ciirrent 
needles (flushed in 1988) and 2-year-old needles (flushed 
in 1986) were separately removed from each sun- and 
shade-shoot. The needles were placed in the assimilation 
chamber keeping the cut end of each fascicle immersed in 
water using a small vinyl tube. Net photosynthetic rates 
at different PFD levels (20-1,300 ^imol/mVs) and dark res- 
piration rates of these two needle groups were measured. 
Air temperature inside the chamber was maintained at 
10 °C throughout the measurement; this temperature 
value was within the range of optimum air temperatiire 
(10-15 °C) in net photosynthesis for P. pumila needles 
(Kajimoto 1990). 



25 



20 - 



15 



o 



< 0 

-5 

-10 
25 



max. 




c 
o 



20 - 



1^ 



10 



£^ 5 
o 

^ 0 




M 



J A 
1988 



0 N 



Figure 1 — Seasonal changes in daily solar irradiance 
(400-1 ,100 nm in wavelength) and air temperatures 
during 1988 measured in a P. pumila stand (2,600 m 
a.s.l.), Kiso Mountain Range. Each point stands for 
the mean value at 1 -week intervals. 



£ 

J3 



V) 

■*-' 
C 
Ct 

V. 

O 

o 

U) 



C 

o 

Q> 

"8 

c 




—I O 



Figure 2 — Seasonal growth patterns of current 
shoots, current needles, and winter buds ob- 
served for sun-exposed P. pumila shoots. Each 
point shows the mean for n = 4 shoots, o = 
shoot length; • = needle length; □ = bud 
length. 



Needle area of each sample was measured as its pro- 
jected area with an area meter (Hayashi, AAC-100), and 
its dry weight was measured after ovendrying at 85 °C. 
Specific leaf area (SLA; cm7g) was determined as the ra- 
tio of needle area to dry weight. Net assimilation rates 
obtained in all measurements were converted to the val- 
ues at a COg concentration of 380 ppm. 

RESULTS 

Climatic Conditions 

Daily solar radiation at the P. pumila stand was higher 
in June and gradually decreased until October in 1988 
(fig. 1). Mean air temperature was about 12 °C during 
August and fell below 0 °C after mid-October. Monthly 
precipitation was 212 mm in July and 167 mm in August 
1988, according to the data obtained at Senjoujiki Station 
(2,623 m a.s.l.) in the central part of Kiso Mountain Range 
(Nagano Meteorological Observatory 1988). 

Growth Patterns of Shoots and 
Needles 

The current shoots elongated rapidly from June to July 
(fig. 2). The current needles began to flush in early July 
and elongated rapidly during August. In late May, the 
winter buds, which would flush in the next season, were 
visible. The length of winter bud increased gradually un- 
til late October. 

Specific leaf area (SLA) of the current needles, both 
sun- and shade-needles, was quite large in early August 
when the needles began to grow larger, and sharply de- 
creased by mid-September (fig. 3). SLA of 2-year-old 



94 



Cur -Shade 




Figure 3 — Seasonal changes in specific leaf area 
(SLA) of the P. pumila needles used for photosyn- 
thesis measurements. □ = current, sun-needles; 
■ = current, shade-needles; O = 2-year-old, sun- 
needles;* = 2-year-old, shade-needles. 



needles became slightly larger in August than in other 
months. The shade-needles, both current and 2-year-old, 
showed larger SLA values throughout the season than 
sun-needles. 




Cur.- Sun 



/ 



Cur-Shod e 
2 yr-Sun 



2yr-Shade 



1000 1200 



Photosynthetic photon flux density 
(pmol nn-2s~i ) 

Figure A — Examples of the relationships between 
photosynthetic photon flux density (PFD) and net 
photosynthetic rate (PJ at 10 °C of P. pumila 
needles on September 15, 1988. □ = current, sun- 
needles; ■ = current, shade-needles; 0= 2-year-old, 
sun-needles; • = 2-year-old, shade-needles. 



PFD 



Light Response of Net Photosynthesis 

Light response of photosynthesis was examined based 
on the relationships between photon flux density (PFD) 
and net photosynthetic rates (PJ at 10 °C (fig. 4). Light 
saturation points of were different between sun- and 
shade-needles, and also between needles of different age. 
However, all sampled needles were completely light satu- 
rated in at PFD value below 1,000 |imol/m7s and 10 °C. 
A maximum net photos5nnthetic rate (P^^) was defined as 
P^ value at 1,000 |imol/mVs and 10 °C, and an approxi- 
mate light-saturation point in net photosynthesis (PFDg^) 
was also determined as PFD value that gave 90 percent 
ofP . 

max 

Table 1 shows that PFDgg of 2-year-old needles ranged 
from 180 to 500 |imol/mVs in sun-needles and 180 to 
440 |imol/m7s in shade-needles. The values were rela- 
tively lower than those of the current needles, where 
PFDgo was 440-580 [imol/mVs in sun-needles and 300-500 
\imo]/mVs in shade-needles. Both current and 2-year-old 
needles had highest PFDgg values in August compared to 
the other months. 

Seasonal Change in Photosynthetic 
Activity 

Figure 5 shows that P^^ of current needles were larger 
in sun-needles than in shade-needles for each period. For 
2-year-old needles, there was a remarkable difference in 

between sun- and shade-needles in July and August. 
The values of sun- and shade-needles both became higher 
in August than in other months. P values of current 

^ max 

needles increased between late August and mid-September. 

Dark respiration rates at 10 °C (R) of current needles 
were larger in sun-needles than in shade-needles, while 
there were no differences in 2-year-old sun- and shade- 
needles (fig. 5). Current needles showed larger R values 
than 2-year-old needles throughout the season. 



Table 1 — Approximate light-saturation points (PFDgg) in the net 
photosynthetic rates (at 1 0 °C) of P. pumila needles in 
1988 

Light-saturation points (PFD^Y 



Date 


Current needles 


2-year-old-needles 


Sun 


Shade 


Sun 


Shade 


May 27 






420 


440 


July 2 






340 


340 


July 20 






300 


300 


August 4 


500 


500 


500 


380 


August 24 


580 


460 


480 


380 


September 1 5 


440 


360 


180 


180 


October 20 


500 


300 


240 


220 



'PFDgg was defined as the value of photosynthetic photon flux density 
which gave 90 percent of a maximum net photosynthetic rate {P^^J- Each 
PFDg^ value was determined using the light-photosynthetic curve, as shown in 
figure 4. 



95 



«» 12 

«A 

* ft, 

c o 
w o 6 

o e 

Q. 

<U _ £ ^ 
0 

<u 

■4-^ MIC 

a V 1.3 



1 - 
0.5- 
0 



- 


/ ^ ^V-:- Cur-Sun 


2yr-Sun 

\ 




- 2yr-Shade 


■N 0 


- 


Cur-Shade 
— ' ' ' ' 


M J 


J A S 0 



2yr-Shade 



2yr-Sun 




M 



0 



1988 



Figure 5 — Seasonal changes in net photosyn- 
thetic rates {P^J at 1 ,000 iimol/m^/s and 
10 °C, and dark respiration rates (fl) at 10 °C 
of P. pumila needles during 1988. □ = current, 
sun-needles; ■ = current, shade-needles; 
O = 2-year-old, sun-needles; • = 2-year-old, 
shade-needles. 



Needle longevity of P. pumila, 4 years or more, corre- 
sponds to that of P. mugo and P. contorta (Benecke and 
Havranek 1980). 

Climate Factors Affecting 
Photosynthetic Production 

Seasonal differences in photosynthetic activity between 
new and old needles have been reported for some north- 
em or subalpine evergreen conifers (Fry and Phillips 
1977; Teskey and others 1984). Likewise, photos5Tithetic 
activity (P^^J in the 2-year-old needles of P. pumila be- 
came higher between late July and early August, while 
that of current needles did so in late August and mid- 
September (fig. 5). A similar seasonal trend in P^^ was 
found for the current and older needles (1 to 4 years old) 
of P. pumila in the study plot in 1987 (Kajimoto 1990). 
These results suggest that photosynthetic production of 
the P. pumila canopy is primarily active during August 
and September. 

Alpine conifer trees rarely achieve their potential maxi- 
mum photosynthetic activities under natural environmen- 
tal conditions, although the main limiting factors in photo- 
sythesis, such as irradiance and air and soil temperatures, 
seem to be different depending on species and site condi- 
tions (De Lucia and Smith 1987; Hasler 1982; Tiirner and 
others 1983). Mean air temperatures between late July 
and mid-September at the research plot (fig. 1) corre- 
sponded to optimiun air temperatm-e (10-15 °C) for net 
photosynthesis of P. pumila, both for current and old 



DISCUSSION 

Growth Characteristics of Needles 

Needles of P. pumila began to elongate in early July 
and almost finished growth in late August (fig. 2). The 
period of needle growth is 1 or 2 months later than those 
of other low-altitude pines, for example, P. densiflora 
(Tanaka and others 1976) and P. thunbergii (Nagatsu 
1987), growing naturally in central Japan. After needle 
growth had terminated, the current needles of P. pumila 
showed gradual decline in specific leaf area (SLA) (fig. 3) 
and maintained higher dark respiration rates than 
2-year-old needles (fig. 5). This indicates that the current 
needles continued to accumulate internal carbohydrate or 
nutrient until October. 

In the study area, there was little local variation in sea- 
sonal growth patterns in the shoots and needles of P. 
pumila (Kajimoto 1989b). However, the final lengths of 
current shoots and needles became smaller with increas- 
ing altitude, as reported for other high-altitude pines, 
such as P. cembra (Baig and Tranquillini 1976), and P. 
mugo and P. contorta (Benecke and Havranek 1980). 

The current needles of P. pumila in the research plot 
accounted for 9 percent of total foliage biomass in late 
July and 27 percent in late August (Kajimoto 1989a). 
This indicates that 70 percent or more of the pine canopy 
consisted of older needles throughout the growing season. 



^ August 

^ 30- 



u 
c 

Of 
D 
CT 
<1> 



> 

"a 
cr 



20 
10 



^ '0 




10 




September 

30 
20 
10 
0 




20 



Air temperature CO 



Figure 6 — Relative frequency distributions of 
daytime (8 a.m. to 4 p.m.) mean air tempera- 
ture at each 1-hour interval in August (above) 
and September (below) 1 988. Air tempera- 
tures were recorded at the P. pumila stand 
(2,600 m a.s.l.), Kiso Mountain Range. 



96 



30- 



o 
c 

u 
cr 



20 - 
10 - 



August 




























1000 



2000 



> 



40 



5 301- 
(T 20h 



September 



10 ^ 



0 *^^^^°"'f" 



0 1000 2000 

PFD ( jjnnol m'h'' ) 

Figure 7 — Relative frequency distributions of 
daytime (8 a.m. to 4 p.m.) photosyntiietic photon 
flux density (PFD) at each 1 -hour interval in 
August (above) and September (below) 1 988. 
PFD was measured above the canopy surface of 
P. pumila stand (2,600 m a.s.l.), Kiso Mountain 
Range. 

needles (Kajimoto 1990). Figure 6 shows patterns of rela- 
tive frequency distribution of daytime mean air tempera- 
ture (8 a.m. to 4 p.m.) at 1-hour intervals; temperature 
values between 10 and 16 °C made up about 70 percent of 
the sum total in both August and September. With re- 
gard to irradiance conditions, proportions of PFD values 
above 600 |j.mol/m7s, when the pine needles were photo- 
synthetically light saturated (table 1), were 53 percent in 
August and 36 percent in September (fig. 7). The tem- 
perature and light regimes indicate that net assimilation 
of P. pumila needles under natural conditions is limited 
mainly by lower levels of solar irradiance. Reduction in 
available PFD from August (53 percent) to September 
(36 percent) is likely to impede net photosynthesis of the 
current needles, since photosynthetic activities (P^^) of 
the current needles were the highest in September (fig. 5). 

It is concluded from the present study that P. pumila 
needles are not very efficient in photosynthesis diiring the 
first growing season. The pine needles survive for four or 
more growing seasons and recover their photosynthetic 
abilities in each summer; P of the oldest needles 

' max 

(4 years old) was about half of that of 1-year-old needles, 
although P^^ decreased with needle age (Kajimoto 1990). 
Consequently, a large proportion of annual net carbon 
gain in the P. pumila tree is likely to depend on net as- 
similation by the old needles. 

Knowledge of physiological properties of P. pumila, 
especially water relations (Ando and Kawasaki 1991), is 
still limited compared to other subalpine pine species. As 
to seasonal photosynthetic production of P. pumila, more 
information on its photosynthetic reponses to environ- 
mental factors, particularly based on field measurement, 
is needed. 



ACKNOWLEDGMENTS 

I thank Dr. K. Hozumi for his suggestions and en- 
couragement during this study, and Drs. T. Koike, 
M. Ishizuka, A. Osawa, Y. Chiba, and N. Kurachi for their 
useful comments on this manuscript. Comments by Drs. 
F.-K. Holtmeier, W. C. Schmidt, and R. Hasler are ac- 
knowledged. Thanks also go to Professors T. Horiuchi 
and S. Ito, Shinshu University, and to the staff of Shinshu 
University Experimental Forest and of Nagoya University 
Forest for their assistance during the field siu"vey. 

REFERENCES 

Ando, H.; Kawasaki, K. 1991. Growth of Pinus pumila in 
relation to the site conditions on a slope. Transactions 
of the 102d Meeting, Japanese Forestry Society: 
451-452. [In Japanese]. 

Baig, M. N.; Tranquillini, W. 1976. Studies on upper tim- 
ber line: morphology and anatomy of Norway spruce 
(Picea abies) and stone pine (Pinus cembra) needles 
from various habitat conditions. Canadian Journal of 
Botany. 54: 1622-1632. 

Benecke, V.; Havranek, W. M. 1980. Phenological growth 
characteristics of trees with increasing altitude, 
Craigieburn Range, New Zealand. In: Benecke, U.; 
Davis, M. R., eds. Mountain environments and subal- 
pine tree growth. Tech. Pap. 70. New Zealand Forest 
Service, Forest Research Institute: 155-174. 

De Lucia, E. H.; Smith, W. K. 1987. Air and soil tempera- 
ture limitations on photosynthesis in Engelmann 
spruce during summer. Canadian Journal of Forest 
Research. 17: 527-533. 

Fry, D. J.; Phillips, I. D. J. 1977. Photosynthesis of coni- 
fers in relation to annual growth cycles and dry matter 
production. II. Seasonal photosynthetic capacity and 
mesophyll ultrastructure in Abies grandis, Picea 
sitchensis, Tsuga heterophylla and Larix leptolepis in 
S.W. England. Physiologia Plantarum. 40: 300-306. 

Hadley, J. L.; Smith, W. K. 1987. Influence of krummholz 
mat microclimate on needle physiology and survival. 
Oecologia. 73: 82-90. 

Hasler, R. 1982. Net photosjTithesis and transpiration of 
Pinus montana on east and north facing slopes at al- 
pine timberline. Oecologia. 54: 14-22. 

Holtmeier, F.-K. 1973. Geoecological aspects of timber- 
lines in northern and central Europe. Arctic and Alpine 
Research. 5(3): 45-54. 

Holtmeier, F.-K. 1981. What does the term "krummholz" 
really mean? Observations with special reference to 
the Alps and the Colorado Front Range. Mountain 
Research and Development. 1(3/4): 253-260. 

Kajimoto, T. 1989a. Aboveground biomass and litterfall of 
Pinus pumila scrubs growing on the Kiso Mountain 
Range in central Japan. Ecological Research. 4: 55-69. 

Kajimoto, T. 1989b. Current growth of Pinus pumila. 
Transactions of the 37th Meeting, Chubu Branch, 
Japanese Forestry Society: 55-58. [In Japanese]. 

Kajimoto, T. 1990. Photosynthesis and respiration of 
Pinus pumila needles in relation to needle age and 
season. Ecological Research. 5: 333-340. 



97 



Kajimoto, T. 1992. Dynamics and dry matter production 
of belowground woody organs of Pinus pumila trees 
growing on the Kiso Mountain Range in central Japan. 
Ecological Research. 7: 333-339. 

Larcher, W. 1975. Physiological plant ecology. Heidelberg: 
Springer. 252 p. 

Nagano Meteorological Observatory. 1988. Annual report 
on weather conditions of Nagano prefecture in 1987. 
Nagano Branch Office of Japanese Meteorological Soci- 
ety. 137 p. [In Japanese]. 

Nagatsu, M. 1987. Correlation between the growth of 
needles and shoots in a Pinus thunbergii yoimg tree. 
Transactions of the 98th Meeting Japanese Forestry 
Society: 417-418. [In Japanese]. 

Okitsu, S.; Ito, K. 1984. Vegetation dynamics of the Sibe- 
rian dwarf pine {Pinus pumila Regel) in the Taisetsu 
Mountain Range, Hokkaido, Japan. Vegetatio. 58: 
105-113. 

Okitsu, S.; Ito, K. 1989. Conditions for the development of 

the Pinus pumila zone of Hokkaido, northern Japan. 

Vegetatio. 84: 127-132. 
Sakai, A.; Larcher, W. 1987. Frost survival of plants. Ecol. 

Stud. 62. New York: Springer. 321 p. 
Shidei, T. 1963. Productivity of Haimatsu {Pinus pumila) 

community growing in alpine zone of Tateyam-Range. 



Journal of the Japanese Forestry Society. 45: 169-173. 
[In Japanese with English simimary]. 

Tanaka, H.; Oohata, S.; Akai, T. 1976. Elongation and 
shoot forms in foreign pines. Report, Kyoto University 
of Forestry. 11: 38-49. [In Japanese]. 

Teskey, R. O.; Grier, C. C; Hinckley, T. M. 1984. Changes 
in photosynthesis and water relations with age and sea- 
son in Abies amabilis. Canadian Journal of Forest 
Research. 14: 77-84. 

Tranquillini, W. 1979. Physiological ecology of the alpine 
timberline. Ecol. Stud. 31. New York: Springer. 131 p. 

Turner, H.; Hasler, R.; Schonenberger, W. 1983. Contrast- 
ing microenvironments and their effects on carbon up- 
take and allocation by young conifers near alpine 
treeline in Switzerland. In: Wareing, R. H., ed. Carbon 
uptake and allocation in subalpine ecosystems as a key 
to management: Proceedings of lUFRO workshop. 1982 
August 2-3. Corvallis, OR: Oregon State University, 
Forest Research Laboratory: 22-30. 

Wardle, P. 1977. Japanese timberlines and some geo- 
graphic comparisons. Arctic and Alpine Research. 9: 
249-258. 



98 



HEIGHT GROWTH IN CEMBRAN PINE 
AS A FACTOR OF AIR TEMPERATURE 

Herbert Kronfuss 



Abstract — The study treats height growth in cembran pine 
iPinus cembra) in high-elevation afforestation at an altitude 
of 1,800 m with growth patterns analyzed as a factor of air tem- 
perature. It is assumed that growth behavior follows an endog- 
enous rhythm possibly deriving from the genotj^je. The relation- 
ship between growth pattern and air temperature is demon- 
strated for defined periods and phases of height growth. From 
the results it becomes obvious that height growth and thus the 
relative rate of growth are primarily dependent on temperature. 
Data from long-term monitoring series were used to determine 
the ampUtudes between the growth curves for early and for late 
culmination. A comparison of the ampUtudes for cembran pine 
with spruce and larch shows cembran pine — with the highest 
amplitudes — to be the most sensitive to temperature. Cembran 
pine thus makes most efficient use of warm weather periods for 
increment production in short periods of time. 



The cembran pine (Pinus cembra) occupies the highest 
stands of all trees in the Alps and as such is ideally suited 
for a study of height growth related to air temperature. 
The fact that cembran pine is not thought to have differ- 
ent ecotypes adapted to various altitudes also permits 
growth characteristics of this species to be defined in spe- 
cific terms. 

STUDY SITE 

The study site was located 30 km southwest of Innsbruck, 
Tyrol (Austria), in the Stubai Alps near Haggen in the 
Sellrain Valley. For location, geology, and climate details, 
see figures 1 and 2, and table 1. The results relate to a 
continuous experimental stand of cembran pine at an alti- 
tude of 1,800 m. The stand was established in 1970 when 
the plot was afforested with 4-year-old plants. 

MONITORING 

Monitoring was performed over a period of 8 years 
for a group of 40 trees. Height growth in cembran pine 
was measured at 5-day intervals (pentades) and the rate 
of growth correlated with mean air temperature of the 
pentade. Individual height increments were expressed 
as a percentage of annual cumvdative grov^^h to permit 
the growth patterns to be compared -with, changes in 
temperatxire. 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Herbert Kronfuss is Forester, Forestry Research Institute of Austria, 
Rennweg 1/29, A-6020 Innsbruck. 



Growth patterns are represented in the form of relative 
growth-rate c;irves and relative cumulative-growth curves. 
This not only permits direct comparisons to be made be- 
tween annual increments but also, given a long enough 
monitoring period, permits a characteristic increment pat- 
tern (type) to be derived for each species (Burger 1926; 
Larcher 1980; Week 1955). 

HEIGHT GROWTH 

From 1978 to 1985 increment growth lasted an average 
of 108 days, vnth a 20-day variance range. 

To study the influence of temperature on the growth 
pattern during the period of extension growth, the rela- 
tive grovrth-rate curve was correlated with the tempera- 
ture curve, with the mean height increments from con- 
secutive pentades expressed as a percentage of total 
height growth. The relative growth-rate curve clearly 
illustrates the pattern of height grov^h (fig. 3). 

Figure 3 shows how temperature fluctuations during 
the period of extension growth influence the rate of height 
growth. With regard to the correlation between the 
height-growth behavior and temperattire patterns, the 
following points would seem to be of interest. 

Dming growing season, the temperature curve itself 
is characterized by two pronotmced dips for the periods 
May 20 to 25 and Jtme 15 to 20, a phenomenon well 
knovm to the local rural communities as the "Ice Saints" 
and "Sheep's Chill." At the beginning of April, mean tem- 
perattire reaches freezing point on the long-term average. 
The rise in temperature continues imtil mid-May, reach- 
ing an average of approximately 7 °C, before dropping 
back to 5 °C for the pentade occupying the first of the two 




Figure 1 — Geographical location of the site. 



99 



HAGGEN i.S.dSOO m) a) 3,1° b) 920 mm 
c) 10 d) 1975 - 1984 



mm 
200 

100 




10 



I I I I I 



I I I I I I I 



I I I I I 



M 



M 



a) Mean annual temperature 

b) Mean annual precipitation 

c) Number of years of data collection for temperature and 
precipitation 

d) Period of data collection for temperature and 
precipitation: 1975-1984 

e) Mean monthly precipitation 

f) Mean monthly temperature 

g) Months with average daily minimum below 0°C 

h) Months with absolute minimum temperatures below 0°C 

Figure 2 — Data for Haggen Climate Diagram 
Haggen I.S., Field Station 1, 1,800 m, SSW. 



cold spells. During the time of the "Ice Saints," most of 
the cembran pines are at the beginning of the period of ex- 
tension growth, with a growth rate of only 2 to 3 percent 
of relative growth per pentade. That corresponds to about 
1.5 cm. On the average, approximately 7 percent of cumu- 
lative growth is achieved by this point in time. For height 
grovi^h in cembran pines in this location, the period be- 
tween the "Ice Saints" and "Sheep's Chill" is of special 
importance as it includes the first major peak in the tem- 
perature curve, on or around May 31 on average. 

With a brief delay, this temperature peak stimulates 
a response in the cembran pine in terms of height growth, 
and after one- and one-half pentades, the biggest absolute 
increment is produced, accounting for about 35 percent of 



total growth on a long-term average. Following this peak 
in increment production, the typical pattern of height 
growth in the cembran pine is one of declining rates of 
growth in spite of rising temperatures. This is shown in 
figure 3 with the two curves moving in opposite directions 
following the start of the "Sheep's Chill" period aroimd 
June 20. 

Not even an increase in mean pentade temperature 
beyond the 10 °C mark around July 5, finally peaking 
at 13 °C, leads to a further spontaneous spate of growth. 



Table 1 — Location of the study: HAGGEN afforestation site in the 
Sellrain Valley near St. Sigmund, Tyrol 



Geographical coordinates: 
Height above sea-level: 
Exposition and angle of slope: 



47°13'N. ir06'E. 
1,715-1,950 m 
SSW, 30° 



Bedrock: Foliated gneiss (biotite granite gneiss, granodiorite 
gneiss) and mica schist. 

Vegetation at 

afforestation: Heath (callunetum, nardetum alpigenum) 
Soils: Humus horizons developed into a brown earth type 

through grazing and haymaking. More or less 

podsolic brown soils. 
In previous centuries a heavily grazed avalanche slope. 

CLIMATIC DATA are based on a 10-year monitoring period (1975- 
1984) 

- Sunshine hours during the vegetation period (mid-May to 

September) 

a) theoretical asatronomical maximum 2,005 hours 

b) theoretical local maximum 1 ,393 hours 
horizon loss 30.5 percent 

- Air temperature 

a) mean annual temperature 3.1 °C 

b) during vegetation period 8.4 °C 

- Precipitation 

a) mean annual precipitation (3 m aboveground) 920 mm 

b) during vegetation period (3 m aboveground) 533 mm 

c) precipitation at ground level 650 mm 
(= approx. 22 percent higher than at 

3 m aboveground) 

- Precipitation probability (percent) 

a) June maximum 55 percent 

b) December minimum 27 percent 

- Mean precipitation distribution (mm) 

a) maximum in July and September 8.4 mm 

b) minimum in February 3.6 mm 

c) month with highest precipitation (July) 141 mm 

d) month with lowest precipitation (February) 33 mm 

- Evaporation during the vegetation period 

a) potential evaporation (2.5 m aboveground 566 mm 
with "piche") (= 4.1 l/m^ per day) 

b) actual evapotranspiration (lysimeter with 236 mm 
ground vegetation (= 1 .7 l/m^ per day) 

c) actual hydrological budget 412 mm 
(= 3 l/m^ per day) 

The meteorological summer semester accounts for 65 percent of 
annual precipitation (598 mm), and the meteorological winter 
semester 35 percent (322 mm). 



100 



Mean pentad values for relative growth rate 
and air temperature (°C) 



Mean relative growth rate (%) 




January February March April May June July August September October November December 

Figure 3 — Comparison between the curves of relative growth rate and air 
temperature, 8-year series (1978 to 1985). 



The minor fluctuations in the growth curve on the order 
of 1 to 3 percent can be correlated with temperature fluc- 
tuations of 10 to 13 °C. 

The main features of the curves in figure 3 can be de- 
scribed as follows: The grovvi;h curve for cembran pine 
peaks (maximum growth rate per pentade is reached) 
on June 5 to 10 on average. The temperature curves are 
characterized by two peaks, the first between the two pro- 
nounced dips (May 31 to June 5) and the second between 
July 25 and July 31. It is interesting to note that height 
increments between July 5 and the end of the period of 
extension growth — and that is a good 46 days — total 
only 20 percent of cumulative grov\i;h, even though the 
temperature curve passes 10 °C on July 5 and remains 
more or less at that level until mid-September. 

The above picture, based on the mean figures calculated 
from monitored data, illustrates in bold strokes the corre- 
lation between the temperature and height-growth patterns. 

The relative grov^i:h-rate curves shown in figure 4 for 
the 8 years from 1978 to 1985 provide an overview of 
growth patterns in cembran pine as a factor of air tem- 
perature for the individual years. One of the more strik- 
ing features of these curves is the variation in the onset 
of culmination (equals greatest increment production per 
pentade). On this basis we can distinguish between an- 
nual growth curves with early culmination (1979, 1981, 
1982, 1983), one year with late culmination (1984), and 
years with twin or intermediate peaks (1978, 1980, 1985; 
see also table 2). 

A comparison of the temperature curve in figure 3 with. 
the growth-rate curves in figure 4 reveals two clearly 
delineated time periods in which culmination (peak 
growth rate per pentade) occurs. The first is the period 
between the "Ice Saints" and "Sheep's Chill," and the sec- 
ond is a subsequent period of 30 days ending approxi- 
mately with the pentade of July 15 to 20. These two peri- 
ods together total 57 days (equals 53 percent of the mean 









1978 


















...z 


■ 1 


. . . . 1 


!^^^7^ ^ 




10 20 31 10 20 30 10 20 31 10 20 31 

May June July August 



W 20 31 10 20 30 10 20 31 10 20 31 









— »« 


















J. 




■ 


^^~.yrr>^ 



10 20 31 10 20 30 10 20 31 10 20 31 









1980 


















■ ■1 








W 20 31 10 20 30 10 20 31 10 20 31 









1981 






















■A^ 



« 20 31 10 20 30 10 20 31 10 20 31 









1983 




















,,l 








10 20 31 10 20 30 10 20 31 10 20 31 




10 20 31 10 20 30 10 20 31 10 20 31 









1986 



























10 20 31 10 20 30 10 20 31 10 20 31 



Figure 4 — Cembran pine relative height-growth 
curves (percent). 



101 



Table 2 — Air temperature in °C and percentage growth in cembran 
pine' 



J n I I 



Year 


TS 


T 


TS 


T 


HG 


HG 






°C- 






- - Percent - - 
















Early culmination 








1979 


245 


9.1 


298 


9.9 


71 


22 


1981 


242 


8.9 


230 


7.7 


61 


19 


1982 


246 


9.1 


368 


12.3 


52 


O IT 

35 


1983 


203 


7.5 


359 


12.0 


46 


38 


Average 


234 


8.7 


314 


10.5 


58 


28 






Intermediate culmination 










8.2 


241 


8.0 


38 


52 


1980 


195 


7.2 


205 


6.8 


26 


47 


1985 


191 


7.1 


284 


9.5 


31 


50 


Average 


203 


7.5 


243 


8.1 


32 


50 






Late culmination 








1984 


169 


6.3 


284 


9.5 


17 


58 



'I = period from the "Ice Saints" to "Sheep's Chill"; II = period from "Sheep's 
Chill" to 20 July; TS = temperature sum; T = mean temperature; HG = height 
growth; period I lasts for 27 days; period II lasts for 30 days. 




10 20 31 10 20 30 10 20 31 10 20 31 

May June July August 



Figure 5 — Cembran pine relative height growth 
(percent). 



period of extension growth) and account for an average 
of 82 percent of annual shoot length. 

There is a significant correlation between percentage 
growth within the two main growth periods and tempera- 
ture sum, v^dth peak growth located in the period that ex- 
hibits the highest temperature sum in absolute terms, 



except that beyond an average temperature sum of 230 °C 
for the first period a further increase in temperature for 
the second period does not trigger higher growth rates in 
that period. The highest pentade growth rate always oc- 
curs in the period with the highest temperature Bum and 
coincides with or follows the pentade with maximum 
pentade temperature (see table 2). 

From table 2 it can be seen that in the case of early cul- 
mination the average temperature sum in the first period 
was 234 °C (8.7 °C/27 days). During that period, an aver- 
age of 58 percent of extension growth was measured (with 
scatter between 46 percent and 71 percent). In those 
cases the cembran pines achieved an average of only 28 
percent relative growth in the second period in spite of the 
rise in the average temperature sum to 314 °C (10.5 °C/30 
days). 

The year of late culmination, 1984, is also instructive. 
The cold weather in period I of that year produced a tem- 
perature sum of only 169 °C (6.3 °C/27 days), which is 
reflected in a correspondingly low relative growth rate 
of 17 percent. Period II, however, was characterized 
by a significant rise in temperature, producing a tem- 
perature sum of 284 °C (9.5 °C/30 days). The relative 
growth rate in that period (fig. 5) was 58 percent, which 
is identical with the figure for period I in the years of 
early culmination. 

At the end of July in the late culmination year, 
the leaders had reached 90 percent of total extension 
growth compEired with 98 percent for the years of early 
culmination. 

This analysis of the height-growth patterns of cembran 
pine over an 8-year period of observation (1978 to 1985) 
shows that the cembran pine is ontogenetically equipped 
to make efficient use of available warmth for increment 
production so as to achieve ctdmination as early as pos- 
sible. Up to that point, air temperature is the dominant 
factor in determining the relative growth rate, £ind after- 
ward it continues to function as the "engine" that main- 
tains the extension process. 

PHASES OF HEIGHT GROWTH 

The following analysis of annual height growth by help 
of the growth curve (Kronfuss 1985) will illustrate the 
somewhat abstract correlation between growth rate and 
temperature with the aid of further figures. For this pur- 
pose, height growth can be divided into the following 
three phases (fig. 6): 

a. Initial phase (0 to 25 percent of annual extension). 

b. Main phase (25 to 75 percent of annued extension). 

c. Final phase (75 to 100 percent of annual extension). 

Table 3 provides a separate assessment of the correla- 
tion for each of these three phases using a regression line, 
correlation coefficient, and regression coefficient to ex- 
press the relationship between mean daily temperature 
(T/d) and average height growth per day (HG/d) (fig. 6, 
table 3). 

Whereas the correlation coefficient for the initial phase 
(r = 0.6) suggests a correlation vnth temperature, this is 
no longer the case in the final phase (r = 0.42). The main 



102 



Correlation between mean daily air temperature and height increment 
per day HAGGEN i.S., 1800 m 

Average height increment per day (mm) 





4- 


r = 0.89 k=^^!£A 






^ Mam phase 












Initial phase 


r=.0.42 /f.O.II 




• 


• • 


r=o.60 ;f=o.is ■ 


• •• 

1 i_ 


* " * Final phase 

_ 1 ■ 


S 6 7 8 


9 10 


It 12 13 14 



Air temperature ("C) 



Correlation between mean daily air temperature and growth processes 
HAGGEN i.S.. 1800 m 

Duration of phases (days) 




Air temperature (°C) 

Figure 6 — Cembran pine correlations between 
mean daily air temperatures, height increment, and 
duration of growth phases: period 1 978 to 1 985. 



growth phase, on the other hand, is largely determined by 
temperature, as can be seen from the almost stretched 
curve (fig. 6 at top) and a correlation coefficient of r = 0.89. 

The regression coefficients {h), as the expression of aver- 
age daily growth per vinit of temperature, show that, for 
the same temperature rise, height growth in the main 



phase is 4.5 times greater than in the initial phase and 
6.2 times greater than in the final phase. In the initial 
phase, with an average period of extension growth of 32 
days, the average daily increment is 1.6 mm. In the 
main growth phase the period of extension growth is re- 
duced to an average of 26 days of an HG/d of 4 mm, while 
the figure for the subsequent 50 days of the final phase is 
only 1 mm. 

The duration of extension growth also correlates with 
temperature and is shorter with higher temperature rise. 
The influence of temperature on the period of extension 
growth can be seen from the correlation coefficients, with 
a close correlation with temperature expressed by the co- 
efficients r = 0.82 for the initial phase and r = 0.84 for 
the main growth period, decreasing in the final phase 
to r = 0.59. This low correlation coefficient shows that 
reduced height growth in the final phase is no longer due 
exclusively to the influence of temperature, but to the 
tailing away of an endogenous pattern. With regard to 
the temperature-dominated phases of extension growth, 
(the initial and main phases) an increase in mean tem- 
perature by 1 °C reduces the extension growth period by 
4 to 5 days. Also, in the final phase we see again that the 
pattern derives from an autonomous endogenous rhythm 
that determines the potential period of extension growth 
in this phase for the species involved. 

GROWTH CURVE VARIANTS 

As can be seen from the shapes of the relative growth- 
rate curves in figure 4, annual patterns of height growth 
vary considerably for the same species in the same loca- 
tions. Eight years of observation offer enough data for 
graphic representation of maximum amplitudes between 
the growth curves for the earliest and those for the latest 
culminations and for a comparison with spruce and larch 
(fig. 7). This illustrates not only the time sequences with 
regard to flushing but also the cxirrent range of fluctua- 
tion for these tree species. 



Table 3 — Correlations between mean daily temperature (t7d) and height growth per day (HG/d) and between mean dally 
temperature (t7d) and the period of extension growth days (d) 



Year 




Initial phase^ 






Main phase^ 






Final phase^ 






t°/d 


HG/d 


d 


t°/d 


HG/d 


d 


t°/d 


HG/d 


d 


1978 


7.9 


1.8 


24 


6.9 


2.6 


33 


10.4 


0.9 


48 


1979 


9.6 


1.8 


22 


8.8 


4.7 


17 


9.9 


.6 


71 


1989 


5.4 


.9 


43 


6.6 


2.5 


32 


12.0 


1.0 


42 


1981 


6.5 


1.7 


25 


.4 


4.7 


18 


8.9 


.7 


59 


1982 


8.2 


1.6 


28 


9.3 


4.1 


22 


11.7 


.9 


52 


1983 


6.1 


1.8 


33 


8.8 


4.6 


26 


13.1 


1.5 


40 


1984 


5.2 


1.3 


52 


9.7 


5.5 


25 


9.8 


1.6 


42 


1985 


7.5 


2.1 


29 


7.2 


3.5 


35 


11.2 


1.4 


43 


Average 


7.1 


1.6 


32 


8.5 


4.0 


26 


10.9 


1.1 


50 



'Initial phase: correlation coefficient r= 0.604; regression coefficient /( = 0.147; point of intersection b= 0.587; standard deviation s= 0.326. 
^Main phase: r = 0.889; k = 0.681 ; b = -1 .471 ; s = 1 .000. 
^Final phase: r = 0.426; /c = 0. 1 1 7; b = -0. 1 95; s = 0.353. 



103 



Cumulative growth (%) / 




^ - -.--^ 


— ^ V/- 

/ 0 / 
/ cr y / 




/-/" /• 

/ '''' 




tj~r" 




r 1 1 1 


1 1 L 1 1 1 J 1 1 1 J 1 1 1 



1 10 20 30 10 20 31 10 20 30 10 20 31 10 20 31 10 20 30 
April Sept. 

Figure 7 — Characteristic growth curve ampli- 
tudes for cembran pine, spruce, and European 
larch: afforestation "Haggen" in 1 ,800 m a.s.l. 



Apart from the differences in the beginning and end of 
the period of extension growth, cembran pine is the most 
sensitive to temperature of the three species, as the high 
amplitude shows. It can be concluded that the cembran 
pine makes the most efficient use of warm weather condi- 
tions for increment production. In the initial phase of 



extension growth, cembran pine is already well ahead of 
the other species in terms of growth. In the main growth 
phase, larch is also surpassed by the cembran pine, while 
spruce comes very close to it. 

For the main growth period, the two extremes of the 
growth curves for spruce and larch are almost parallel. 
This shows that increment production per pentade is 
more balanced between warm and cold conditions than 
in the case of cembran pine, which is much more sensitive 
to temperature fluctuations. 

REFERENCES 

Burger, H. 1926. Untersuchungen iiber das Hohen- 
wachstum verschiedener Holzarten. Mitt. Schweiz. 
Centralanst. Forstl. Vers, wesen; XIV/I. 

Kronfuss, H. 1985. Die Zuwachsleistimg einer Hochlagen- 
aufforstung mit Zirbe (Pinus cembra) auf einem 
Siidhang, in Abhaangigkeit von der Seehohe. lUFRO 
Meeting Proceedings. Davos, Switzerland. 

Larcher, W. 1980. Okologie der Pflanzen auf 
physiologischer Grundlage. Ulmer, Stuttgart. 

Week, J. 1955. Forstliche Zuwachs und Ertragskimde. 
Neimiann, Berlin. 



104 



SURVIVAL AND GROWTH OF PLANTED 
CEMBRAN PINES AT THE ALPINE 
TIMBERLINE 



Josef Senn 

Walter Schonenberger 
Ueli Wasem 



Abstract — In an experimental afforestation with cembran pines 
{Pinus cembra L.) at the subalpine timberhne, tree survival was 
mainly determined by the date of disappearance of the snow 
cover in spring. Trees survived well on early snow-free sites and 
poorly on sites with long-lasting snow cover. The major causes of 
mortality were the two parasitic fungi Ascocalyx abietina and 
Phacidium infestans. Tree height, as an index for growth, was 
dependent on the amount of energy available during the vegeta- 
tion season, and therefore negatively correlated with both alti- 
tude above sea level and wind velocity, and positively with global 
radiation. 



In the severe winter of 1951-52 avalanches caused 
major destruction all over the Alps. This initiated a 
multidisciplinary research program to restore the upper 
treeline. The aim of the program was to develop methods 
to alforest treeless avalanche catchments and abandoned 
meadows within the subalpine forest belt in order to es- 
tablish new protection forests. 

Protection forests are an important feature in alpine 
landscapes. Many areas can only be permanently inhab- 
ited as long as these forests remain intact. In some of 
these forests, however, severe problems concerning stand 
stability have become evident. Inadequate plantation 
techniques in the past have frequently led to forests that 
were highly susceptible to natural hazards such as 
storms. On the other hand, locally dense ungulate popu- 
lations prevented successful natiu-al rejuvenation in the 
existing forests. 

Both natural regeneration and artificial afforestation 
generally have become more difficult toward the alpine 
timberline, since all biological processes become slower 
with increasing altitude above sea level. In general, com- 
pared to natiiral invasion of forest trees, planting reduces 
the time needed to establish forests that fulfill their pro- 
tective function. Therefore artificial afforestation may be 
highly desirable. Poor financial resources, however, may 
Umit planting to the most promising sites. 

The experimental afforestation at StiUberg was designed 
to obtain information about site conditions that determine 
success or failure of future afforestations, and to be able 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Josef Senn, Walter Schonenberger, and Ueli Wasem are members of the 
High Altitude Afforestation Group at the Swiss Federal Institute for For- 
est, Snow and Landscape Research, CH-8903 Birmensdorf, Switzerland. 



to inform forestry personnel which sites would be most 
promising for successful planting and on which sites 
planting would be fruitless. 

Cembran pine {Pinus cembra L.) is one of the three tim- 
berline species tested at StiUberg. This species, extending 
to the highest altitudes of any trees in the Alps, played an 
important role in the culture of local human populations 
(Holtmeier, these proceedings), produces valuable timber, 
and is esteemed by "green" tourism as an important fea- 
ture of natural landscapes. Nonsustainable exploitation 
in medieval times and pasturalism reduced cembran pine 
to a fraction of its original abundance, although there is 
evidence for natural range expansion in the past 100 years. 
Artificial reproduction and estabhshment of cembran pine, 
on the other hand, pose many problems for practical for- 
estry (Frehner and Schonenberger, these proceedings). 

MATERIAL AND METHODS 

The present study was conducted at StiUberg experi- 
mental area, a highly structiu-ed northeast slope (fig. 1) 
in the Dischma Valley near Davos, Switzerland (lat. 
34°08' S., long. 171°41' E.). The area extends over 9.9 ha 
between 2,080 and 2,230 m above sea level. The highest 
parts reach beyond the altitudinal timberline. Before 
plantation in 1975 various site aspects like slope, expo- 
sure, global radiation (Turner 1966), wind conditions 
(Nageli 1971), date of disappearance of snow cover in 
spring (Rychetnik 1987), ntunber of snow-fi-ee days in 
winter, avalanche fi'equencies, soil types (Blaser 1980), 
and vegetation (mainly dwarf shrub communities) (Kuoch 
1970) were recorded on a very fine scale. Schonenberger 
(1975) described the site in detail. Results from the 
StiUberg experimental afforestation were published by 
Schonenberger and others (1988). Practical aspects for 
high-altitude forestry were included in Schonenberger 
and others (1990). 

Three coniferous tree species were planted in 1975: 
5-year-old cembran pines, 3-year-old mountain pines (P. 
uncinata (Miller) Domin = Pinus montana Miller = Pinus 
mugo Turra ssp. uncinata Domin) the erect form of the 
mugo pine, and 1-year-old European Isirches (Larix de- 
cidua Miller). Before planting, the whole area was di- 
vided into square units of 3.5 by 3.5 m (fig. 1). In each 
square unit 25 trees of one species were planted. The 
number of trees per imit may have been reduced by ad- 
verse local soil conditions like barren rock. The squares 
contained alternatively cembran pines, mountain pines, 
and larches. Altogether some 90,000 trees were planted 
in 4,052 square units. 



105 




Figure 1 — Topographic model of the Stillberg 
experimental area with the grid of the affores- 
tation units (3.5 by 3.5 m). 



This paper reports the results of survival of the cembran 
pines in comparison with mountain pine and European 
larch in the whole study area, and of height variation and 
mortality causes in a subsample. Survival of each tree 
was recorded annually as from 1976 (number of square 
units with cembran pines = 1,351, with mountain pines = 
1,350, and with larches = 1,351). Mean tree height, gen- 
eral constitution, type and degree of abiotic and biotic 
damage, and causes of mortality were determined annu- 
ally in 228 square units containing cembran pines, in 226 
square imits containing mountain pines, and in 226 
square units containing larches. Twenty-eight qualitative 
"types" of damage and potential causes of death could be 
attributed in the field. 

Correlation analysis was used to test for interactions 
between various site parameters and tree performance. 
The relative importance of site parameters was evaluated 
by stepwise regression analyses (SAS 1985) for survival 
and tree height. The survival rates were arcsine trans- 
formed before performing stepwise regression analysis. 

RESULTS 
Survival 

During the first 3 years after planting, survival was 
highest in cembran pine compared to mountain pine and 
larch, but in 1991 only 25.3 percent of the planted cembran 
pines survived, compared to 37.8 percent in mountain 
pine and 73.1 percent in larch (fig. 2). 

Stepwise regression of survival rate in a square unit 
(number of trees alive in 1991/number of trees planted in 
1975) of cembran pine revealed that the date of disappear- 
ance of snow cover in spring and the altitude above sea 
level were the most important factors, explaining some 
46 percent and 8 percent of total variation, respectively. 
In our study area the survival rate was high on the early 
snow-free and lower sites. Survival decreased with in- 
creasing altitude, and almost no trees survived on sites 
where snow lasted on 10-year average longer than June 10. 



Avalanche frequency, slope inclination, global radiation, 
and the number of snow-free days in winter also had sig- 
nificant effects on survival, but each of these factors con- 
tributed less than 1 percent to the total variation. 

Mortality Causes 

Of the 28 recorded causes for mortality, Ascocalyx 
able- tina (Lagerb.) (= Gremmeniella abietina [Lagerb.] 
Morelet), a parasitic fungus, was the major factor, killing 
some 52.3 percent of the planted cembran pines. A second 
fungus, Phacidium infestans Karst., the snow blight, 
killed another 11.8 percent of the trees. Some 4.8 percent 
of the trees died or disappeared without any detectable 
reason. Undetermined fungi killed 1.3 percent and com- 
petition with herbaceous vegetation some 0.6 percent of 
the planted trees. Herbivores caused no losses. 

We found a significantly positive correlation between the 
proportion of cembran pines killed by Ascocalyx abietina 
and the date of disappearance of snow cover in spring 
(fig. 3). Square units that were snow free before May 12 
suffered only slightly (13.9 percent of the trees killed), 
whereas AscocaZyx killed 90.3 percent of the trees in units 
that were covered by snow later than June 10. Occur- 
rence oi Ascocalyx was also positively correlated with 
altitude {r = 0.55, P < 0.0001, n = 228), the fungus kilHng 
an increasing proportion of yoimg trees with increasing 
altitude. 

Between the proportion of cembran pines killed by Pha- 
cidium infestans and the date of disappearance of snow 
cover in spring, however, we found a negative correlation 
(r = -0.42, P < 0.001, n = 228). This means that the high- 
est proportion of trees was killed in squares that were 




75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 



Year 

Figure 2— Survivorship curves of the three 
tree species, cembran pine (Pinus cembra), 
mountain pine (P. uncinata), and European 
larch {Larix decidua), planted in 1975 in the 
Stillberg experimental area. 



106 



1.0 



p 

E o 

O 0) 

a 
o 



0.8 • 



0.6 



0.4 



0.2 



• . . • • • 



• • • I • 



•i 

•t . 



• • • 

t'..' «. 



120 130 140 150 160 170 

Date of Disappearance of Snow Cover (day) 

Figure 3 — Relationship between the propor- 
tion of cembran pines killed in a square unit by 
Ascocalyx abietina (= number of trees killed by 
Ascocalyx since 1 976/number of trees planted 
in 1975) and the date of disappearance of snow 
cover in spring. Day 1 30 = May 1 0; day 1 60 = 
June 9. Regression line: y = -3.0637 + 0.0251 x, 
r=0.8^,P« 0.0001, n = 228. 



snow free early, whereas the importance of Phacidium 
as a mortaHty cause decreased with increasing duration 
of snow cover. Similarly the ratio of trees killed by Pha- 
cidium decreased with increasing altitude {r = —0.26, 
P<0.0001, n = 228). 

The occurrences of the two parasitic fungi were signifi- 
cantly negatively correlated within square units (r = -0.59, 
P < 0.0001, n = 228). If one of the two fungus species was 
common in one square, the other was rare. 



DISCUSSION 
Survival 

Duration of snow cover in spring was found to be the 
main limiting factor for survival of young cembran pines 
in the upper subalpine forest belt. On sites where snow 
cover lasted until after Jime 10 cembran pines were al- 
most completely excluded. The second evergreen Pinus 
species, the mountain pine, was similarly vulnerable to 
long-lasting snow cover. The European larch, a deciduous 
conifer, was much more tolerant. In this species 30 per- 
cent of the trees siu^ved on sites snow covered until after 
June 10. Further, survival of cembran pine was signifi- 
cantly affected by altitude, although this factor was clearly 
less important than duration of snow cover in spring. In- 
terestingly, the impact of site factors such as global radia- 
tion, number of snow-free days, slope inclination, and ava- 
lanche frequency on tree survival was negligible compared 
to the first two factors. Avalanche frequency, however, 
will certainly become more important in the near future. 
Young trees with flexible stems generally survive ava- 
lanche impacts without significant damage. But with in- 
creasing height and stem diameter the trees are becoming 
increasingly vulnerable to stem breakage caused by snow 
movements, a tendency that has become evident only dur- 
ing the last years. 

Mortality Causes 

In our study area the two parasitic fungi Ascocalyx 
abietina and Phacidium infestans were the major mortal- 
ity factors for cembran pines, and these fungi still kill sig- 
nificant numbers of young trees every year. 



Tree Height 

Mean tree height in 1991 was lowest in cembran pine 
(58.8 cm) compared to moxmtain pine (72.7 cm) and larch 
(59.8 cm), although initial mean height was largest in 
cembran pine (fig. 4). 

Variation in height was considerable, depending on 
local site conditions. On sunny southeast exposed sites, 
average height of cembran pines was above 80 cm, whereas 
on shady sites mean height was around 45 cm. Stepwise 
regression revealed that tree height was mainly affected 
by altitude, global radiation, and wind velocity, these fac- 
tors contributing 21 percent, 15 percent, and 10 percent 
to the total variation. Duration of snow cover in spring 
explained 3 percent of the variation in height. In our 
study area the cembran pines were high on sites at low 
altitudes, receiving high amounts of radiation and being 
protected from strong winds. Slope inclination, avalanche 
frequency, and number of snow-free days in winter had no 
significant impact on tree height. 



E 

o 



c 

(0 
0) 



80 



70 



60 



SO 



X 



30 



20 



10 































1 








-♦■ - 


- C€ 

- M< 

- La 


>mb 
}unt 
rch 


'an pine 
ain pine 














r' 


r' 


— 






-«-•• 










r' 
























, 


✓ 


r' 


r'' 








b-' 



















✓ 




...I 

p: — 


r' 


— yt 
(' 


1 1 



































— 










r-- 




r-- 


r 


..-1 

r' 




















r'' 


r 



















1 .- 










■ 



75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 

Year 

Figure 4 — Annual mean height of the three tree 
species, cembran pine {Pinus cembra), mountain 
pine (P. uncinata), and European larch (Larix de- 
cidua), planted in 1975 in the Stillberg experimen- 
tal area. 



107 



Ascocalyx infestations first become evident in spring 
in the dying buds. Later in the season the shoots die back 
along the branches. The fungus may eventually kill a 
young cembran pine after 2 or 3 years. The close relation- 
ship between the date of disappearance of snow cover in 
spring and the occurrence Ascocalyx (fig. 3) indicates 
that the fungus preferentially attacks weakened plants. 
On the sites with long-lasting snow cover, the vegetation 
season is relatively short, and thick layers of raw himius 
have accumulated because of the low temperatures. 
Therefore, they are poor sites for tree growth. Similarly, 
Kurkela (1984) found that the trees most heavily affected 
hy Ascocalyx had the lowest growth rate before the epi- 
demic. He explained the differences in tree growth by the 
different site fertility. In our study the hypothesis that 
Ascocalyx preferentially attacked weakened trees was fur- 
ther supported by the finding that the proportion of trees 
killed by the fungus significantly increased with increas- 
ing altitude (toward the altitudinal tree line). In our study 
area Ascocalyx attacked mountain pines, too, where it also 
caused considerable losses. 

In contrast to the situation in Ascocalyx, the proportion 
of cembran pines killed by Phacidium was negatively cor- 
related with the date of disappearance of snow cover and 
with the altitude. According to this correlation the trees 
growing at lower altitudes and on sites that were snow 
free early were killed more frequently by Phacidium than 
the trees growing at higher altitudes and on sites with 
long-lasting snow cover. Young trees that looked healthy 
in one year were often found dead in the following spring. 
Interestingly, Roll-Hansen and others (1992) reported 
from Scandinavia that the most vigorously growing plants 
were most susceptible to Phacidium infestations, whereas 
poorly developing plants were most resistant. 

The fact that Phacidium killed the highest proportion 
of trees at early snow-free sites seems contradictory to 
the biology of the fungus, that exclusively attacks parts 
of trees that are covered by snow (Roll-Hansen 1989). 
Branches destroyed by this fungus are clearly visible after 
disappearance of snow cover and mark the prevalent snow 
depth. Our study area contains some older cembran pines 
growing on ridges (at the sites that are snow free early). 
These trees originating from caches of the nutcracker 
{Nucifraga caryocatactes) host Phacidium, which is non- 
lethal for the large trees that extend considerably beyond 
the snow cover. These trees may act as sources for local 
Phacidium outbreaks that eventually kill the young trees 
completely covered by snow. Since this fungus specifically 
attacks cembran pine and does not occur on mountain 
pine (but see Roll-Hansen 1989), it may not spread as eas- 
ily over larger areas as does Ascocalyx, which uses moun- 
tain pine as an alternative host. Therefore, even heavier 
outbreaks of Phacidium remain locally restricted around 
mature cembran pines. Together with the fact that Pha- 
cidium preferentially attacks the most vigorous trees, this 
may explain the seemingly paradoxical negative correla- 
tion between duration of snow cover in spring and occur- 
rence o{ Phacidium. 

On sites with high avalanche frequencies the young 
trees were partially excluded, not by stem breakage or 
uprooting, but by competitive interactions with herba- 
ceous vegetation. The moving snow carried rich mineral 



soil leading to a lush herbaceous vegetation that out- 
competed the slowly growing young cembran pines. Ava- 
lanches, however, become a more serious problem for the 
surviving trees as they increase in height and stem diam- 
eter. Broken stems were found more frequently in recent 
years. 

Herbivore impact on the survival of young cembran pines 
was relatively unimportant. Herbivores killed no trees, al- 
though they may reduce tree growth. Browsing vertebrates 
like black grouse (Tetrao tetrix) and chamois (Rupicapra 
rupicapra), however, clearly preferred mountain pine and 
larch to cembran pine. Black grouse caused some locally 
restricted damage by browsing buds and first-year needles. 
They obviously preferred to browse on easily accessible 
trees growing on ridges with little snow cover (Streule 
1973). Populations of herbivorous insects like phloem- 
sucking aphids (Cinara cembrae Seitner, see Grbic, these 
proceedings, and Pineus cembrae [Choi.] Amand.) fluctu- 
ated in size but were low in most of the years when com- 
pared to the other two tree species. Although visible dam- 
age by herbivores was relatively rare, much rarer than 
fungal infections, the herbivores may have indirectly af- 
fected the survival of the trees by injuring plant tissue, 
making the trees more susceptible to infestations by fungi. 

Tree Height 

Tree height in cembran pine was clearly dependent on 
the available amount of energy during the vegetation sea- 
son. Altitude above sea level, global radiation, and wind 
velocity explained the largest amount of variation in 
height. We are aware that tree height does not directly 
represent tree growth, since we did not measure annual 
growth. We only recorded tree height in summer, exclud- 
ing the growing shoot of the current year. Tree height, 
which is the result of annual growth minus the losses 
through damage in the apical region, may increase or 
decrease from one year to another. Sixteen years after 
plantation, differences in average tree height among sites, 
however, are primarily the result of variation in growth 
caused by variation in local site conditions. 

The date of disappearance of snow cover in spring, al- 
though significantly related to tree height, explained only 
3 percent of the variation in height in contrast to survival 
where the snow conditions in spring mainly explained the 
observed pattern. 

Slope inclination and avalanche frequency, which may 
be related to stem breakage, had no detectable impact on 
tree height. Further, we found no relationship between 
the number of snow-free days in winter and tree height, 
although trees that were snow free during extended peri- 
ods in winter should have been more vulnerable to herbi- 
vore browsing than trees that were covered by deep snow. 

CONCLUSIONS 

Foresters should carefully select favorable microsites 
when planting cembran pines in high-altitude afforestations. 
Particular attention should be paid to the spatial pattern 
of snowmelt in spring, and no cembran pines should be 
planted at sites with prolonged duration of snow cover. 



108 



REFERENCES 

Blaser, P. 1980. Der Boden als Standortsfaktor bei 
Aufforstiingen in der subalpinen Stufe (Stillberg, 
Davos). Eidgenossische Anstalt fur das forstliche 
Versuchswesen, Mitteilimgen. 56(3): 527-611. 

Kuoch, R. 1970. Die Vegetation auf Stillberg (Dischmatal, 
Kt. Graubiinden). Eidgenossische Anstalt fiir das 
forstliche Versuchswesen, Mitteilungen. 46(4): 329-342. 

Kurkela, T. 1984. The growth of trees affected by Grem- 
meniella abietina. In: Manion, P. D., ed. Scleroderris 
canker of conifers. The Hague: Martinus NijhofiEDr W. 
Junk Pubhshers: 177-180. 

Nageh, W. 1971. Der Wind als Standortsfaktor bei 
Aufforstungen in der subalpinen Stufe (Stillbergalp 
im Dischmatal, Kanton Graubiinden). Eidgenossische 
Anstalt fiir das forstliche Versuchswesen, Mitteilungen. 
47(2): 33-147. 

Roll-Hansen, F. 1989. Phacidium infestans: a literature 
review. European Journal of Forest Pathology. 19: 
237-250. 

Roll-Hansen, F.; Roll-Hansen, H.; Skroppa, T. 1992. 
Gremmeniella abietina, Phacidium infestans, and other 
causes of damage in alpine, young pine plantations in 
Norway. European Journal of Forest Pathology. 22: 
77-94. 

Rychetnik, J. 1987. Snow cover disappearance as influ- 
enced by site conditions, snow distribution, and ava- 
lanche activity. In: Proceedings of the international 
symposiima on avalanche formation, movement and 
effects; Davos 1986. lAHS-Pubhcation Nr. 162. 

SAS. 1985. SAS user's guide: statistics. Gary, NC: SAS 
Institute. 



Schonenberger, W. 1975. Standorteinfliisse auf 
Versuchsaufforstimgen an der alpinen Waldgrenze 
(Stillberg, Davos). Eidgenossische Anstalt fiir das 
forstliche Versuchswesen, Mitteilungen. 51(4): 357-428. 

Schonenberger, W. 1985. Performance of a high altitude 
afforestation under various site conditions. In: T\irner, 
H.; Tranquillini, W., eds. Establishment and tending of 
subalpine forest: research and management: Proceed- 
ings of the 3rd lUFRO workshop P 1.07-00, 1984. 
Eidgenossische Anstalt fiir das forstUche Versuchswesen, 
Berichte. 270: 233-240. 

Schonenberger, W.; Frey, W., eds. 1988. Untersuchimgen 
zur Okologie und Technik der Hochlagenaufforstung; 
Forschungsergebnisse aus dem Lawinenanrissgebiet 
Stillberg. Schweizerische Zeitschrift fiir Forstwesen. 
139: 735-820. 

Schonenberger, W.; Frey, W.; Leuenberger, F. 1990. 
Okologie und Technik der Aufforstung im Gebirge - 
Anregimgen fiir die Praxis. Eidgenossische Anstalt fiir 
das forstliche Versuchswesen, Berichte Nr. 325. 58 p. 
[Version franfaise: Ecologie et technique des afforesta- 
tions en montagne - Suggestions a I'usage des praticiens. 
Traduction du rapport No. 325.] [Versione italiana: 
Ecologia del rimboschimento di montagna e pratica 
della tecnica di piantagione. Traduzione del rapporto 
No. 325.] 

Streule, A. 1973. Schaden in Gebirgsaufforstungen durch 
das Birkhuhn (Lyrurus tetrix). Biindnerwald. 28(8): 
249-254. 

Tiimer, H. 1966. Die globale Hangbestrahlung als 
Standortsfaktor in der subalpine St\ife. Eidgenossische 
Anstalt fiir das forstliche Versuchswesen, Mitteilungen. 
42(3): 111-168. 



109 



Influence of 
Environmental Factors 




International Workshop 
St. Moritz 1 992 

s ^ ' 



110 



CHANGES OF SWISS STONE PINE 
APHID LIFE CYCLE, DENSITY, AND 
POPULATION STRUCTURE IN HIGH- 
ALTITUDE SWISS STONE PINE 
AFFORESTATION 

Mihailo Grbic 



Abstract — The order of form appearance, structure, and number 
of individuals in colonies of Cinara cembrae on Swiss stone pine 
{Pinus cembra) used for afforestation in the Dischma Valley were 
strongly related to the slope aspect. The abnormality was par- 
ticularly apparent on north slopes. North slope colonies 
showed incomplete life cycles. All insects seen there were 
virginoparae and their larvae. Colonies were small in number 
and were developed from winged virginoparae from other areas. 



The Alps are of great importance to Switzerland as an 
area for hviman habitation and as a health and holiday re- 
sort. Afforestation in high altitudes is costly and involves 
high risk, as many failures occur. Because of the slow de- 
velopment of such afforestation it is difficult to under- 
stand and to determine the causes and processes that lead 
to a failure and then to suggest measures to improve 
success. 

Environmental factors of the upper timberline area are 
totally different from those of the forest zones. These 
harsh conditions affect the specific growth and develop- 
ment of planted species; for example, low survival rate 
and low annual increment. The surviving plants, depend- 
ing on the biological characteristics of the species, are 
more or less open to insect and fungal attack. These same 
high-altitude factors (for example, low temperature, short 
growing period, and short day photoperiod) tend to change 
the life cycle, density, and population structure of insects. 
This is particularly true for aphids due to their polymor- 
phic nature, and their ability to alter the nimaber of gen- 
erations per season. 

STUDY AREAS AND METHODS 

These investigations were carried out in field plots of 
the Stillberg research area and the Lucksalp comparative 
afforestation area, in the timberline area of the Dischma 
Valley (Canton Graubunden). The period of the study 
was between late May and the middle of September, 
which is the time of greatest insect activity. 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Mihailo Grbic is Teacher at the University of Belgrade, Faculty of For- 
estry, Kneza Viseslava 1, YU-11030 Belgrade, Yugoslavia. 



Stillberg Hes at 2,080 to 2,230 m above sea level (a.s.l.), 
with a northeast aspect, and 30° to 45° slope. The slope is 
divided by spurs, and as a result there are three different 
aspects (north-, northeast-, and east-facing slopes). On 
the opposite side of Dischma Valley lies Lucksalp at 
2,200 m a.s.l., with a southwest aspect. 

The subjects of investigation were 19-year-old Swiss 
stone pine (Pinus cembra L.) and the Swiss stone pine 
aphid (Cinara cembrae Seithner), the most frequently ob- 
served insect pest on that afforestation species. The trees 
were planted 70 cm apart in a grid pattern in approxi- 
mately 4,000 square plots (3.5 by 3.5 m). The Swiss stone 
pine stocks were alternated from plot to plot with two 
other afforestation species, 17-year-old Swiss mountain 
pine (Pinus mugo ssp. uncinata [Mill.] Domin.), and 
15-year-old European larch (Larix decidua Mill.), each 
plot containing 25 trees (Shonenberger 1985). 

The samples of Swiss stone pine aphid were identified 
by comparing them to recent accounts from various parts 
of Europe (Carter and Maslen 1982; Eastop 1972; Pintera 
1966; Stroyan 1955, 1960). The structure of antennae, 
rostrum, abdomen, and hind legs of collected material 
were compared with these references. 

The life cycle was observed in colonies feeding on par- 
ticular sample trees. On Lucksalp, 30 trees (5 trees x 
6 plots) were regtdarly examined, and 50 (5 trees x 
10 plots) were examined on Stillberg. Observations were 
made to establish the relationship between the intensity 
of infestation and site aspects. 

The trees were examined every 10 days to determine 
density of pest population and structvire of colonies. 

MORPHOLOGY 

The morphological characteristics of specimens found 
on Stillberg and Lucksalp (tables 1 and 2) are similar to 
the central European description (Pintera 1966). The 
chronological series of morphs is: fundatrix, aptera 
virginopara, alata virginopara, ovipara, and aptera and 
alata males. 

Compared with the apterae of the succeeding genera- 
tions, fundatrix shows "fundatrix facies" characteristics 
(Lees 1961). With the Swiss stone pine aphid, the anten- 
nae and legs are relatively shorter in comparison to the 
body. The processus terminalis is shorter with fewer 
rhinaria. Features similar to the fundatrix facies that 
were observed by Hille Ris Lambers (1955), Lees (1961), 
and Stroyan (1960) on nonfimdatrix morphs of some 



111 



Table 1 — Biometric data for different morphs of Cinara cembrae from Lucksalp 



Biometric data 



Aptarae 
virginoparae 



Alatea 
virginoparae 



Morph 
Apterae 
male 



Alatae 
male 



Sexual 
female 



Body length (mm) 

Length of antennal segments 
II 
IV 
V 

VI (base + proces. termlnalls) 

Length of rostal segments (^) 
IV 
V 

Hind tarsus segments (^i) 
I 

basal diameter 
dorsal length 
ventral length 

II 

length 

Hind tibia length (mm) 

Length of longest hair on 
III ant. segment 
hind tibia 

VIII abdom. tergite 

Number of hairs on 
II ant. segment 
VI ant. segment 
base 

proces. term, 
ultimate rostral segment 

Number of hairs on 
subgenital plate 
VIII abdom. tergite 

Number of secondary rhinaria on 
ant. segments 

III 

IV 

VI 



3.05-3.52 



680-720 
300-340 
400-410 
167-180 + 
75-83 

218-238 
100-112 



74- 78 
95-105 

155-182 

385-420 
3.20-3.30 

75- 97 
75-120 

110-165 

9-15 

9- 13 
4-5 
6-8 

10- 12 
4-6 



0-2 
0-2 
2-3 



3.00-3.10 



635-650 
275-290 
320-345 
150-160 + 
59-66 

170-182 
85-88 



57-62 
62-66 
122-130 

330-350 
2.90-3.05 

88-92 
70-105 
185-200 

6-7 

10-12 
3-4 
8 

10 



9-11 
2-3 
5 



2.50-3.30 



830-880 
335-400 
430-500 
190-198 + 
75-90 

230-250 
90-110 



60-70 
98-110 
150-180 

340-390 
1.85-3.00 

75-120 
75-110 
100-112 

10-15 

10-12 
5 

10-12 

10-12 
7-11 



10-13 
7-10 
4-5 



2.40-3.00 



845-910 
360-430 
440-500 
109-198 + 
85-95 

220-230 
105-110 



60-65 
85-105 
150-180 

385-400 
2.90-3.40 

115-140 
140-180 
90-120 

7- 14 

8- 10 

5- 6 
8-10 

6- 9 
6-9 



55-83 
10-13 
5 



3.80-4.70 



830-890 
360-385 
480-500 
192-212 + 
75-80 

245-250 
106-118 



68-88 
105-120 
187-200 

370-420 
3.70-3.78 

95-123 
112-145 
120-138 

10-16 

8- 13 
5-6 

9- 11 

9-13 
7-8 



2-3 
1-2 
4-7 



aphids that live at low temperatures were not recorded in 
these study areas. 

Morphological differences between apterous and alatae 
virginoparae are found not only in the presence or ab- 
sence of the wings, but also in the following characteris- 
tics. In alatae: 

• Segmentation between head and prothorEix is more 
conspicuous. 

• Compound eyes are bigger, v(ath many facets. 

• Antennae bear five times as many rhinaria. 

• The body is shorter, and the mesothoracic terga con- 
sists of a compound notum and postnotum. 

• The scutum occupies the central major part of the 
nottun and is divided in two mesothoracic lobes, which are 
hardly developed. 



The virginoparae of every generation are morphologi- 
cally uniform, and under field conditions it is impossible 
to find out how many generations have developed. Also, 
the length of time required for passing through four in- 
stars, from birth to adult, is variable and dependent on 
two external factors (food quality and temperature) and 
two internal factors (birth weight and whether the morph 
is winged or unwdnged) (Dixon 1987). This is the reason 
for the presence of different generations of the same form 
occurring simultaneously in one colony. 

As is known, Swiss stone pine aphids produce both 
alatae and apterous males. Both forms were found in the 
research areas, but with a preponderance of the wingless 
form. Both forms of males are readily recognized by their 
sclerotized genital structure. The body, especially abdo- 
men, is smaller and more slender, and the antennae bear 
more rhinaria. 



112 



Table 2 — Biometric data for different morphs of Cinara cembrae from Stillberg 



Morph 





Aptarae 


Alatea 


Apterae 


Alatae 


Sexual 


Biometric data 


virginoparae 


virginoparae 


male 


male 


female 


Body length (mm) 


2.90-3.70 


2.90-3.90 


2.50-3.10 


2.40-3.00 


3.90-4.70 


Length of antennal segments 












II 
II 


540-800 


^ Art rt C rt 

630-650 


820-800 


840-930 


800-860 


IV 


260-355 


250-280 


320-400 


360-440 


350-400 


V 


340-450 


www OOU 


tOU-wuU 


ARn-t^ 1 n 

HOU 0 1 u 


u-ouu 


VI (base + proces. termlnalis) 


170-188-1- 


153-167 + 


183-210 + 


192-200 + 


175-222 + 




80-85 


58-68 


75-95 


88-95 


75-85 


Length of rostal segments (n) 












IV 


rtOQ OCA 

238-250 


1 62-1 72 


235-250 


220-225 


235-250 


V 


100-108 


83-87 


95-102 


100-105 


105-118 


Hind tarsus segments (n) 
1 












h^^^sl dif^mptAr 


75-78 


WW w9 




DU-DO 


DO- 1 UO 


dorsal length 


88-108 


60-62 


95-1 02 


88-1 07 


105-125 


ventral length 

II 


158-175 


122-127 


150-178 


150-170 


185-230 


length 


365-385 


300-340 


335-395 


380-400 


420-450 


Hind tibia length (mm) 


2.40-3.39 


2.82-2.98 


1 .78-2.98 


2.95-3.37 


3.70-3.88 


Length of longest hair on 












III ant. segment 


60-00 


85-88 


78-125 


1 1 2-1 42 


95-130 


hind tibia 


60-105 


75-1 10 


78-1 12 


140-175 


1 12-150 


VIII aooom. tergite 


168-220 


195-210 


100-115 


88-130 


112-138 


Number of hairs on 












II ant. segment 


11-15 


6-7 


10-16 


7-15 


10-18 


VI ant. segment 












base 


8-12 


9-13 


10-13 


8-1 1 


8-16 


proces. term. 


4-5 


3-4 


5-6 


5 


4-6 


ultimate rostral segment 


8-9 


7-9 


9-12 


8-11 


8-11 


Number of hairs on 












subgenital plate 


9-11 


6 


9-12 


6-9 


9-14 


VIII abdom. tergite 


4-6 




7-12 


6-9 


6-9 


Number of secondary rhinaria on 












ant. segments 












III 


0-4 


8-10 


10-15 


50-89 


2-4 


IV 


2-3 


1-3 


7-11 


11-15 


0-2 


VI 


3-4 


6 


4-6 


5-6 


2-7 



Alatae and apterous males and alatae virginoparae pos- 
sess a far larger number of secondary rhinaria than apter- 
ous virginoparae and sexuales females. The following 
numbers of secondary rhinaria on the third antennal seg- 
ment were found: apterous virginoparae 0-4, alatae 
virginoparae 8-11, apterous males 10-15, and alatae 
males 50-89. The number of secondary rhinaria on 
sexuales females was similar to that found in apterous 
virginoparae. 

The function of secondary rhinaria is probably different 
in male and female individuals. The secondary rhinaria 
in males appear to be the main receptors of sex phero- 
mones, while secondary rhinaria of alatae virginoparae 
must have a function other than pheromone detection. 
This function is possibly host selection, as has been shown 
by many authors with other aphid species (Alikhan 1960; 
Chapman and others 1981; and Pospisil 1976). 



Sexuales females resemble the apterous virginoparae, 
but may be somewhat bigger. The tibiae of the hind legs 
are often longer with longer setae also, and numerovis 
pseudosensoria on them. 

LIFE CYCLE 

General — The life cycle of the Swiss stone pine aphid is 
strongly related to site aspects and environmental factors. 
Generally, the Swiss stone pine aphid has a monoecious 
holocyclic type of life cycle. The life cycle starts in May 
v^dth the hatching of fundatrices larvae from eggs laid in 
rows on the needles. In the covirse of svmimer a small 
niunber of generations of parthenogenetic females 
develop. 

Generations of winged parthenogenetic viviparous fe- 
males arise from colonies of apterae later in the summer. 



113 



Sexuales forms appear in the second half of August. This 
early development of sexuales is due to the alpine climate. 
In the low-lying positions where Swiss stone pine has 
been artificially introduced, Swiss stone pine aphid has 
never been found. 

Observed — The life cycle observations in the Dischma 
Valley show the following characteristics: colonies from 
Lucksalp and all slopes of Stillberg, except north-facing 
slopes, show complete life cycles with six morphs. Par- 
allel appearance of two forms of males, observed by 
Seithner (1936), is a unique case in the family Lachnidae. 
The north-slope colonies showed incomplete life cycles. 
All insects seen there were virginoparae apterae or alatae 
and their larvae. Colonies were small in number and 
were developed from winged virginoparae from other 
areas that colonized plants on north-facing slopes from 
other areas (table 3). 

The order of appearance of particular forms on 
Lucksalp was earlier than on the Stillberg research area. 
The first observation, made in the last third of May, 
shows 68 percent unhatched eggs, and all fundatrix lar- 
vae were part of the first two larval instars. At the same 
time on Stillberg 87 to 90 percent of the eggs were 
unhatched. 

In the first third of June the colony structure on 
Lucksalp was: 7 percent of the eggs were unhatched, 
young fundatrices in all four larval instars were found, 
but no adults. A similar, but slightly different, situation 
was recorded on Stillberg. There were 25 to 38 percent 
unhatched eggs, and the colonies contained only the first 
three larval instars. 

During the observations in the second third of June we 
found adult fundatrices on Lucksalp and on the east 
slopes of Stillberg. With the appearance of fundatrices, 
the density of populations increased gradually. 

On the other Stillberg slopes (ENE and NE) fundatrices 
were found 10 days later. Simultaneously we recorded 
adults of apterous virginoparae on Lucksalp. This is a 
time of heavy increases in population, because the multi- 
plication rate of apterous virginoparae is very high, and 
the duration of larval instars very short. 

In an experiment that was conducted under conditions 
of changeable room temperature, one apterous virgino- 
parae was isolated on a twig that was kept moist. Ten 
days later, we recorded six apterous virginoparae of 
a new generation (which started with larviposition); 16 
in the fourth larval instar; and 18 larvae were in yoimger 
stadiums. 

It was presumed that the effect of the multiplication 
rate is not the same in nonisolated field conditions, but 
on some trees the increments of population for 10 days 
increased by a factor of 10 or more. On the other hand, 
some other populations were substantially reduced by the 
action of parasites and predators. 

In the first third of July at Lucksalp a typical popula- 
tion consisted of more than 170 individuals; almost all 
belonged to apterous virginoparae larvae (83 percent) or 
adult insects ( 15 percent). The rest (2 percent) were 
alatoid larvae of alatae virginoparae. The fundatrix gen- 
eration was dead. At Stillberg, populations were smaller 



with most of the population made up of larvae and adults 
of apterous virginoparae; however, fundatrices were also 
found to be present. The fourth larval instar of alatae 
virginoparae was recorded on the east slopes only. The 
proportion of alatoid larvae was 10 percent of the total 
number of the fourth larval instar. 

The first Swiss stone pine aphid was recorded on the 
north slopes of Stillberg in the middle of July. Colonies 
were small and consisted of only a few adult alatae 
virginoparae and young larvae. Few of the colonies were 
without adults. The population structures of the other 
slopes were unchanged in content until the last third of 
August. Only the numbers of individuals were variable, 
depending on weather conditions. 

The rapid decrease in number of individuals at the be- 
ginning of August was caused by low temperatures and 
snow. The greatest changes in population quantity were 
recorded on north slopes. The average population struc- 
ture in the course of July and August was: The majority 
of individuals belonged to various larval instars, and the 
adult part of the population consisted of apterae and 
alatae viviparous parthenogenetic females. Average ratio 
of winged and wingless forms was 1:5.3 with an increas- 
ing tendency to winged forms in colder sites. This ten- 
dency is due to the strategy of the species to avoid expo- 
sure to more extreme climatic conditions (by colonizing 
another host plant of the same species, but in better envi- 
ronmental conditions). Extremes of the ratio were 1:8 and 
1:2. The first was recorded on Lucksalp, and the second 
on Stillberg north slopes. 

The highest population numbers were found in the last 
third of July, after which the numbers decreased rapidly 
as a consequence of low temperatures at the beginning 
of August. Thereafter, the population number increased 
quickly again to a secondary peak at the end of August. 
Occurrence of the last generation was recorded after an- 
other cool weather period at the end of August. As the 
overwintering eggs are produced sexually, the last genera- 
tion of the year must contain males and sexual females, 
instead of the parthenogenetic forms that occur during 
the rest of the year. However, sexuales do not occur un- 
der all investigated conditions. On the north slopes of 
Stillberg, for example, no sexuales forms were recorded. 
On Lucksalp, occurrence of sexuales forms was about 
10 days before Stillberg. 

The first appearance of eggs was observed at the begin- 
ning of September. Oviparous females move away from 
the colony to the top part of current shoots, and lay over- 
wintering eggs in rows on needle clusters. Eggs were 
brown and later black and shiny 48 hours after being laid. 
Average number of eggs in the row along one needle was 
8.2 and the extremes were 16 and 1. Maximum egg num- 
ber per one colony was 316, recorded in the first week of 
September at Lucksalp. The same colony contained eight 
sexuales females. Results of dissection of females show 
12 ripe eggs in the ovariole on average. This was smaller 
in comparison with the embryo numbers obtained by dis- 
section of parthenogenetic females during July, which 
was 30. 



114 



Table 3 — Average number of different morphs of Cinara cembrae on Swiss stone pine sample trees from last third of May to second third of 
September 





Morph 


May 




June 




Julv 






August 




September 


3/3 


1/3 


2/3 


3/3 


1/3 2/3 


3/3 


1/3 


2/3 


3/3 


1/3 


2/3 












Lucksalp (SW slope) 














Eqqs 


5.6 


0.6 
















60.0 


93.0 


Larvae 


2.6 


7.6 


35.0 


103.0 


148.0 178.0 


199.0 


68.0 


82.0 


124.0 


9.0 


11.0 


Fundatrices 






1.2 


1.5 
















Apterae virginoparae 








2.0 


27.0 12.0 


24.0 


8.3 


18.0 


24.0 


7.3 


6.7 


Aiatae virginoparae 










1.5 


2.4 


1.8 


1.5 


2.0 


0.2 




Oviparous females 


















0.1 


4.9 


5.4 


Apterae males 


















0.1 


0.3 


1.4 


Aiatae males 




















0.1 














Stillberg (E slope) 














Eqqs 


7.2 


2.0 


0.2 














3.8 


57.0 


Larvae 


0.8 


4.4 


30.0 


41.0 


58.0 100.0 


110.0 


54.0 


80.0 


104.0 


5.9 


8.1 


Fundatrices 






0.6 


1.8 


1.0 














Apterae virginoparae 










3.6 30.0 


31.0 


6.3 


11.0 


20.0 


2.9 


3.2 


Aiatae virginoparae 












3.0 




5.0 


5.8 




0.7 


Oviparous females 




















0.2 


2.3 


Apterae males 




















0.1 


0.3 


Aiatae males 




















0.1 


0.2 












Stillberg (ENE slope) 














Eqqs 


7.0 


3.0 


1.1 


















Larvae 


0.8 


4.8 


6.7 


34.0 


124.0 140.0 


145.0 


54.0 


69.0 


93.0 


4.6 


8.3 


Fundatrices 








2.4 


0.2 














Apterae virginoparae 










5.0 12.0 


18.0 


3.2 


10.0 


17.0 


3.0 


3.5 


Aiatae virginoparae 












2.6 


1.0 


3.8 


4.7 


1.8 




Oviparous females 




















1.4 


5.2 


Apterae males 




















0.8 




Aiatae males 




















0.2 


0.5 












Stillberg (NE slope) 














Eggs 


6.5 


2.0 


















8.0 


Larvae 


1.0 


5.9 


7.7 


32.0 


40.0 89.0 


118.0 


26.0 


40.0 


58.0 


8.6 


12.0 


Fundatrices 








2.3 


0.8 














Apterae virginoparae 










1.8 10.0 


18.0 


2.3 


7.1 


9.4 


3.8 


9.2 


Aiatae virginoparae 












0.9 


1.0 


1.3 


3.0 


2.6 


2.8 


Oviparous females 




















0.8 


2.0 


Apterae males 




















0.7 


1.5 


Aiatae males 


































Stillberg (N slope) 














Eggs 
























Larvae 










5.8 


43.0 


1.0 


11.0 


18.0 






Fundatrices 
























Apterae virginoparae 












6.2 


1.0 


2.0 


4.8 






Aiatae virginoparae 










1.0 


3.4 




1.6 


1.2 







Oviparous females 
Apterae males 
Aiatae males 



115 



SEASONAL MOVEMENTS 

The most important factors affecting aphid feeding and 
nutrition are the change, in time, of growth and develop- 
ment of the host plant (Klungauf 1987). In conjmiction 
with this change, seasonal movements of colonies were 
observed. 

The fundatrix generation moves from the needles where 
overwintering eggs were laid. Feeding sites of fimda- 
trices, and later generations of virginoparae, in the course 
of Jime and first half of July were on last year's shoots. 
During July, colonies moved to twigs and branches in 
lower parts of the crown (in extreme cases some of the 
colonies were fovmd on stems). This was just after the 
peak of the annual increment of the shoots, when they 
start to become lignified. It seems to be related to de- 
scending sap movement. Conifer lachnids, as phloem 
feeders, exploit food sources in lower parts of the plant 
that become available as the season progresses. 

The alatoid larvae of alatae virginoparae usually move 
from the colony to the tops of the shoots before their final 
moult. Here they moult into alatae adults and fly off to 
colonize similar positions on other plants. 

Simultaneously with the occurrence of sexuales, the 
colonies moved ascendentally. At the end of the season 
the majority of the individuals were on last year's and 
current season's shoots. From this position oviparous 
females moved to lay eggs on needles. 

Additional trials were carried out with colonies isolated 
by polytene covers, which prevented migration from top to 
bottom parts of the host plant. In the beginning, the den- 
sity of isolated colonies was much higher than the density 
of nearby nonisolated colonies. However, early occurrence 
of alatae forms indicated an attempt to change the colo- 
nies' position. 

Later, complete starvation of the population was caused 
by restriction of movement. It was concluded that the 
life cycle of Swiss stone pine aphid is very much synchro- 
nized with the growth and development characteristics of 
their host plant, and in conjunction with this, quality of 
nutrition. 

PEST DAMAGES 

Keen (1938) reported that several species of lachnids 
cause considerable injury to conifers in North America. 
Fumiss and Carolin (1977) described visible signs of in- 
jury, such as yellowing of needles or reduction of growth 
of young trees. According to Seithner (1936), the damage 
to Swiss stone pine is quite iinimportant, but he believed 
that the trees could be weakened by the sucking and that 
secondary pests (for example, Pissodes and Pityophthorus) 
could easily attack them. 

We found no direct visible signs of damage by Swiss 
stone pine aphid except for some discolorations on feeding 
sites. Our observations show that retardation of the 
growth increment by Swiss stone pine aphid, if any, is 
much less significant than retardation due to other envi- 
ronmental factors. 

Indirectly, Swiss stone pine aphid's honeydew excre- 
tion, in the absence of heavy rain, accumulates on the 
surface of the shoots and forms an ideal substrate for 



saprophytic sooty mould fungi to develop. A black cover- 
ing of hyphae and spores can reduce the quality of trees 
that are grown in parks, but for the Stillberg and 
Lucksalp research areas it is not so important. On the 
other hand, honey derived from honeydew, particularly 
from conifer lachnids, is highly valued and of considerable 
economic significance. Fossel (1971) found that the Swiss 
stone pine aphid produced copious honeydew in the movm- 
tainous districts of Austria. 

CONCLUSIONS 

Information gathered during this study has allowed the 
following conclusions to be drawn: 

• Swiss stone pine aphid was the most common insect 
pest on Swiss stone pine used for afforestation in the re- 
search areas of the Dischma Valley. 

• The morphological characteristics of specimens and 
the chronological series of morphs found on Stillberg and 
Lucksalp are similar to the conventional descriptions. 
However, for some site aspects of the Dischma Valley the 
life cycle of Swiss stone pine aphid was incomplete. The 
life cycle abnormality was observed on the north slopes of 
Stillberg. 

• The order of appearance of particular insect forms, 
structure, and number of individuals in colonies also were 
strongly related to site aspects. This is particiilarly true 
for the temperature effect that caused later development 
and smaller sizes of colonies on the north slopes. 

• Swiss stone pine aphid was observed to cause only 
unimportant discolorations on the feeding sites. The 
damage to the trees is much less significant than that pro- 
duced by the normally harsh environmental conditions of 
the upper timberline area. 

When looked at in isolation, the damage produced by 
the pest observed in this study cannot be considered seri- 
ous. However, it is only one link in a chain of factors det- 
rimental to the growth of afforestation plants. How the 
insect interacts with other factors harmful to tree growth 
(for example, fungi) is as yet unknown, and therefore the 
damage caused by colonies of Swiss stone pine aphid can- 
not be accurately assessed. 

Until it is known how detrimental the aphid is, the im- 
portance of control measures or predators and parasites 
also cannot be estimated. 

Clearly, 6 months investigation is too short for con- 
clusive results. However, the information gained has 
provided an important first step in imderstanding the 
relationship between the insect pest, Swiss stone pine 
aphid, and the host plant, Swiss stone pine, in high- 
altitude afforestation. 

REFERENCES 

Alikhan, M. A. 1960. The experimental study of chemotac- 
tic basis of host specificity in a phytophagous insect. 
Aphis fabae Scop. (Aphididae: Homoptera). Annales 
Univ. Mariae Curie - Sklodowska, Sectio C. 15: 117-158. 

Carter, C. I.; Maslen, N. R. 1982. Conifer lachnids in 
Britain. Alice Holt Lodge, Farnham, Surrey, Engleind: 
Forestry Commission Research Station. 



116 



Chapman, R. F.; Bemays, E. A.; Simpson, S. J. 1981. 
Attraction and repulsion of the aphid, Cavariella 
aegopodii, by plant odors. Journal of Chemical Ecology. 
7: 881-888. 

Dixon, A. F. G. 1987. Parthenogenetic reproduction and 
the rate of increase in aphids. In: Aphids: their biology, 
natural enemies, and control. 2A: 269-287. 

Eastop, V. F. 1972. A taxonomic review of the species 
of Cinara Curtis occurring in Britain (Hemiptera: 
Aphididae). Bulletin of the British Museimi, Natiiral 
History (Entomology). 27: 103-186. 

Fossel, A. 1971. New observation on Cinara cembrae 
(Homoptera, Lachnidae). Annales Zoologici. 28: 353-365. 

Fumiss, R. L.; Carolin, V. M. 1977. Western forest in- 
sects. Misc. Publ. 1339. Washington, DC: U.S. Depart- 
ment of Agriculture, Forest Service. 654 p. 

Hille Ris Lambers, D. 1955. Hemiptera 2. Aphididae. 
Zoology of Iceland. 3(52a): 1-29. 

Keen, F. P. 1938. Insect enemies of western forests. Misc. 
Publ. 273. Washington, DC: U.S. Department of 
Agriculture. 

Klingauf, F. A. 1987. Feeding, adaptation and excretion. 
In: Aphids: their biology, natural enemies, and control. 
2A: 225-253. 



Lees, A. D. 1961. Clonal polymorphism in aphids. In: In- 
sect polymorphism. Sjmiposia of the Royal Entomologi- 
cal Society of London. 1: 68-79. 

Pintera, A. 1966. Revision of the genus Cinara Curt. 
(Aphidodea, Lachnidae) in Middle Europe. Acta ent. 
bohemoslov. 63: 281-321. 

Pospisil, J. 1976. Sensory, olfactory and visual orientation 
of insects of the tropical region III. Visual and olfactory 
orientation of alate forms of the aphid, Aphis gossypii 
Glover, in Cuba. Academia de Ciencias de Cuba, Serie 
Biologica. 66: 17-23. 

Schonenberger, W. 1984. Performance of a high altitude 
afforestation under various site conditions. Proc. 3d 
lUFRO Workshop: 233-240. 

Seither, M. 1936. Lachnus cembrae an. sp., die Zirben- 
blattlaus. Centralbl. ges. Forstwes. 62: 33-49. 

Stroyan, H. L. G. 1955. Recent additions to the British 
aphid faima. Part II. Transactions of the Royal Entomo- 
logical Society of London. 106: 283-340. 

Stroyan, H. L. G. 1960. Three new subspecies of aphids 
from Iceland (Hem., Hom.). Entomologiske Meddelelser. 
29: 250-265. 



117 



GENETIC CONSEQUENCES AND 
RESEARCH CHALLENGES OF BLISTER 
RUST IN WHITEBARK PINE FORESTS 

Raymond J. Ho^ 
Susan K. Hagle 
Richard G. Krebill 



Abstract — Susceptibility of whitebark pine {Pinus albicaulis) to 
blister rust (caused by Cronartium ribicola) is reviewed. Prog- 
ress is reported on studies that assess the level of susceptibility 
over its entire range and the existence of resistance in various 
stands. Two breeding approaches are discussed: (1) the tradi- 
tional, where trees are selected, tested, then established in seed 
orchards; (2) a natural approach that aids natural processes to 
establish "a natural selection stand." 



In 1910, eastern white pine (Pinus strobus) seedlings 
that had been grown in Europe were planted in the area 
of Point Grey, BC, Canada. Some of these seedlings were 
infected with white pine blister rust (caused by the fungus 
Cronartium ribicola). The disease was finally noticed in 
the fall of 1921, when it was observed on endemic western 
white pine (Pinus monticola). The destructive nature of 
this disease in North American white pines had already 
been documented in Europe (Spaulding 1911). 

What followed was a frantic rush to stop blister rust's 
spread by destroying infected trees, eradicating currants 
and gooseberries (genus Ribes) that are alternate hosts for 
the disease, and using several promising fungicides. But 
this work failed, and the fungus kept spreading. By about 
1960, the fungus had spread throughout most the range 
of whitebark pine (Pinus albicaulis) (fig. 1), which also in- 
cludes most of the range of western white pine and sugar 
pine (Pinus lambertiana). Examples of blister rust on 
whitebark pine are illustrated in figure 2. 

The first infected whitebark pine recorded was in 
the arboretum of the University of British Columbia, 
Vancouver, BC, in 1922 (Bedwell and Childs 1943). The 
first discovery of blister rust on native whitebark pine 
was in 1926 on the Birkenhead River in the Coast Range 
of British Columbia, Canada (Childs and others 1938). 
Intense infection and mortality by blister rust was not far 
behind (Bedwell and Childs 1943). 

The degree of infection was 100 percent in many stands 
in the northern areas, and decreased to the south where 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Raymond J. HofF is Research Plant Geneticist, Intermountain Research 
Station, Forest Service, U.S. Department of Agriculture, Moscow, ID 83843, 
USA. Susan K. Hagle is Pathologist, Northern Region, Forest Service, U.S. 
Department of Agriculture, Missoula, MT 59807, USA. Richard G. Krebill 
is Assistant Station Director and Pathologist, Intermountain Research 
Station, Forest Service, U.S. Department of Agriculture, Ogden, UT 84401, 
U.S.A. 



whitebark pine occurs in higher and drier sites. White- 
bark pines grovnng below about 45° N. lat. in Idaho and 
Montana have much less rust than whitebark of higher 
latitudes. However, in the Cascade-Sierra Nevada Moun- 
tains the rust has spread to about lat. 36° 15' N., where it 



Pinus albicaulis 

BH malor subalpine component 




Figure 1 — Range map of whitebark pine showing 
areas of high, moderate, and little or no blister rust 
incidence. 



118 




has been especially devastating to western white pine 
and sugar pine. Whitebark pine too has been heavily 
impacted in most of these mountain ranges. But no 
infection was found on whitebark pine at Tioga Pass, 
Yosemite National Park, at about 39° N. lat. in 1992. 

The relative susceptibility of the white pines to blister 
rust has been tested by several workers. Bingham 
(1972b) siimmarized these reports (table 1). All reports 
agreed that whitebark pine was most susceptible of the 
tested five-needled pines. Fiu-ther work was done by 
Bedwell and Childs (1943) to compare the relative suscep- 
tibility of whitebark versus western white pine. These 
two pines grow adjacent for much of their ranges, with 
whitebark pine at the high-elevation sites and western 
white pine lower. They often are foimd on the same site 
in the overlap zone. Childs and Bedwell (1948) found that 
susceptibility (in terms of nvunbers of cankers and speed 
of tree mortality) of whitebark pine was several times 
that of western white pine. They concluded that this dif- 
ference was due in part to (1) a higher susceptibility of 
current year whitebark pine needles and (2) longer needle 
retention of whitebark pine (5.3 years for whitebark and 
3.8 years for western white pine). 



Table 1 — Comparison of susceptibility in white pines to blister rust 
by Bingham 1972a or b and a reshuffle by Hoff and others 
1980 



Species Bingham 1972a or b Hoff and others 1980 



Pinus armandii 


1 


4 


P. cembra 


1 


3 


P. aristata 


1 




P. wallichiana 


2 


6 


P. koraiensis 


2 


1 


P. peuce 


3 


5 


P. sibirica 


4 


2 


P. parviflora 


5 


3 


P. strobiformis 


6 


8 


P. strobus 


7 


11 


P. flexilis 


8 


10 


P. monticola 


9 


4 


P. lambertiana 


10 


9 


P. albicaulis 


11 


7 



119 



Bingham (1983) found one western white pine in 10,000 
that was free of bhster rust. The possibihty of finding 
disease-free whitebark pine trees seemed pretty sHm. 

Nonetheless, at sites in northern Idaho and western 
Montana where mortahty is over 90 percent due to blister 
rust, we have found a few individuals with no or just a 
few cankers. Though they certainly are rare (fig. 3). This 
gives us hope that there are resistant trees that can be 
used to reestablish populations in these areas. 

Hofif and others (1980) presented data indicating that 
three phenotypically resistant whitebark pine trees in- 
cluded in a species trial did indeed have resistance to blis- 
ter rust. In this same study, resistant collections of west- 
em white pine and sugar pine were also included. This 
resulted in a reshuffle of the white pines for susceptibility 
to blister rust (table 1). In this comparison, whitebark 
pine moved up along with resistant western white pine 
and resistant sugar pine. However, these same species 
were tested in France by Delatour and Birot (1982), in 
Japan by Yokota (1983), and in Germany by Stephan 
(1985), and the resistance level of the collections of white- 
bark, western white, and sugar pine was much lower 
(tables 2, 3, 4). The only explanation is that there are 
different races of blister rust at these other test sites. 

In spite of this bad news, our plans are to proceed to de- 
velop resistance in whitebark pine, including as many 




Figure 3 — Blister rust-resistant candidate trees at Gisborne Mountain (A) 

120 



genes for resistance as possible. If new races of the fungus 
appear, the new resistant varieties should have adequate 
flexibility to exhibit resistance to the new races. This 
seems justified, since high levels of resistance in western 
white pine and sugar pine are still being observed in blister 
rust resistance tests, even though several races of the fun- 
gus are evident. 

ACTIVE WORK AND PRELIMINARY 
PROGRESS 

Blister Rust Surveys of Stands, Clearcuts, and 
Bums — Onr objectives are to determine the level of blis- 
ter rust damage over the reuige of whitebark pine and to 
locate phenotypically resistant trees. As mentioned, the 
degree of damage is associated with latitude. Phenotypi- 
cally resistant trees have been located in most stands 
where high mortality to blister rust has occurred. We 
also observed a relatively large amount of natural regen- 
eration occurring in clearcuts and bums adjacent to 
stands of whitebark pine that have had high rust-caused 
mortality. Table 5 shows the number of trees per acre 
and the degree of infection by blister rust in four such 
stands. Natural regeneration of whitebark pine occurs 
from unused seed caches of Clark's nutcracker. The 




Burl^e's Peak (B). 



Table 2 — Percent of rust-infected seedlings in bulk collections of 
white pine species (from Delatour and Birot 1 982) 



Geographic 




Number 


Percent 


group 


Species 


of lots 


infected 


European 


P. cembra 


12 


0 




P. peuce 


7 


0-20 


Asian 


P. sibirica 


2 


0 




P. parviflora 


3 


0 




P. koraiensis 


7 


0-9 




P. wallichiana 


10 


7-67 




P. armandii 


1 


58 


American 


r. ailolala 


1 


o 
c. 




P. albicaulis 


5 


40-100 




P. flexilis 


17 


0-100 




P. strobiformis 


9 


40-100 




P. lambertiana 


5 


64-96 




P. strobus 


24 


88-100 




P. monticola 


5 


94-100 



abundance of regeneration was surprising. It immedi- 
ately broxight up questions: At what level of mortality will 
there not be enough seed available for caching by the nut- 
cracker? If the seed did not come from the adjacent high- 
mortality stand, where did it come from? 

Artificial Inoculation With Blister Rust — The objec- 
tive is to determine the level and variety of mechanisms 
that impart resistance in phenotypically resistant trees 
and in stands that have had high mortality by blister 

Table 3 — Results of inoculation in France and Idaho. 

Percent of rust infection on "resistant" crosses 
of P. lambertiana and P. monticola from 
inoculation in France or Idaho (Delatour and 
Birot 1 982) 

Seed lot No. France 
and cross mean Idaho 



Percent 

P. lambertiana 



41 CF^ 


99 


55 


41 BF, 


98 


57 


41 AF, 


52 


25 


P. monticola 






43 BF, 


96 


40 


43 EF, 


93 


30 


43 CF, 


76 


40 


43 A F, 


32 


30 


43 FF, 


73 


65 


43 DF, 


35 


45 


47 A F, 


75 


65 


46HF2 


63 


35 


46EF2 


56 


40 


46BF2 


55 


45 


46FF2 


44 


35 


46G F2 


40 


30 


46DF2 


38 


20 


46 AFj 


37 


30 


46CF2 


24 


15 



Table 4 — Comparison of blister rust susceptibility of some white 
pines from Stephan (1985) 



North America 


Percent 


Eurasia 


Percent 


species 


infection 


species 


infection 


P. albicaulis 


97 


P. armandii 


0 


P. aristata 


66 


P. cembra 


0 


P. balfouriana 


90 


P. koraiensis 


23 


P. flexilis 


98 


P. morrisonicola 


40 


P. lambertiana 


^97 (76) 


P. parviflora 


22 


P. monticola 


99 


P. peuce 


22 


P. strobiformis 


88 


P. pumila 


0 


P. strobus 


100 


P. sibirica 


17 






P. wallichiana 


40 



'Two seedlots were tested; one was not as susceptible as the other. 



rust. A trial test was inoculated in August 1991 using 
methods outlined by Bingham (1972a). This test included 
about 1,000 seedlings from 36 families from 10 populations. 
The inoculation was light; only 26 percent of the seedlings 
developed needle spots. However, there was considerable 
variation among populations (table 6). Even with the light 
inoculation, it was encouraging to observe several seedlings 
that had what we call "premature needle shed" (McDonald 
and Hoff 1970). This is a resistance mechanism in which 
needles with needle spots are shed from 9 to 12 months 
after inoculation and before the rust fungus has invaded 
the stem. This type of resistance has been observed in 
several white pine species (Hoff and others 1980). The 
seedlings were reinoculated in August 1992 and as of 
November 1, 1992, 87 percent had needle spots. 

Adaptive Variation of Whitebark Pine — How far 

can we transfer seed £ind not cause maladaptation in the 
seedlings? The range of whitebark pine extends nearly 
19 degrees of latitude and 20 degrees of longitude, and 
from 1,524 to 3,354 m. In addition, the species grows in 
various ecological situations. It can fill the role of a pio- 
neer and a serai. It can exist for long periods as a domi- 
nant on harsh sites and even in some stands with the cli- 
max species, subalpine fir. Whitebark pine is found in 
106 of 225 habitat/phase types in eastern Idaho-western 
Wyoming, central Idaho, and Montana (Pfister and others 
1977; Steele and others 1981, 1983). 

However, whitebark pine is relatively intolerant to 
shade and highly sensitive to competition (Arno and 
Weaver 1990). Therefore it is frequently restricted to 
tougher sites that prevent or restrict the growth of other 



Table 5 — Number and infection of whitebark pine seedlings in four 
clearcuts 



Name of Seedlings Blister rust 

clearcut per hectare infection 







Percent 


Vermillion Pass 


445 


55 


Upper Coal Creek 


1,544 


4 


Upper Big Creek 


884 


33 


Divide Mountain 


3,929 


52 



121 



Table 6 — Variation in various traits of whitebark pine for 1 0 provenances 



Preliminary population comparisons 











Third-year 


riea 


hXi 


blister 


^Ji* Aif An Q n A 


1 a* 

Lai. 


Long. 


Elevation 


height 


stem 


DUO 


rust 








m 


cm 




— Percent- - 




Cooper Pass 


47-32N 


115-44W 


1,494 


17.1 


22 


9 


22 


Gisborne 


48-21 N 


116-44W 


1,692 


15.0 


10 


4 


24 


Freezeout 


47-01 N 


116-02W 


1,707 


10.6 


3 


0 


5 


Lunch Peak 


48-22N 


116-22W 


1,982 


16.9 


10 


9 


33 


Brundage Mountain 


45-01 N 


116-07W 


2,195 


16.2 


17 


8 


24 


Seven Devils 


45-21 N 


11 6-31 W 


2,302 


12.8 


6 


3 


26 


Saddle Mountain 


45-42N 


113-59W 


2,378 


18.1 


24 


2 


46 


Porphyry Peak 


46-49N 


110-44W 


2,509 


12.0 


0 


13 


33 


Boulder Peak 


41-35N 


1 23-05 W 


2,523 


10.7 


0 


0 




Palmer Mountain 


45-04N 


110-35W 


2,652 


16.3 


0 


22 


41 


Mean 








14.6 


9 


7 


25 



species. Because of the combination of several severe en- 
vironmental factors, such as short growing season, cold 
air, and snow blast (McCaughey and Schmidt 1990), 
whitebark pine forms krummholz stands of shrublike 
trees at or near timberline. At lower elevations, where 
growing conditions are not quite so severe, whitebark pine 
grows in nearly pure stands of trees 30 to 90 feet tall 
(McCaughey and Schmidt 1990). When environmental 
conditions are even less severe, whitebark pine is associ- 
ated with other tree species and takes on a form much 
like lodgepole pine. When growdng conditions are ideal, 
whitebark pine seedlings will grow fairly rapidly (fig. 4). 

With such a wide-ranging and ecologically diverse spe- 
cies one would expect several geographic races. Since 
nutcrackers normally cache the seed at least several him- 
dred meters and in many cases several kilometers from 
its source, it is also easy to specvilate that there are few 
races because genes from several diverse populations are 
being continually mixed. Future tests should clarify this 
puzzle. 

Seed from about 40 stands have been collected. Our 
first area of study will be Idaho and Montana. The test 
will be established with seed from about 100 populations. 
A preliminary test has revealed some interesting data 
(table 6). There is considerable variation among the popu- 
lations in 3-year-old total height (fig. 4). The tallest seed- 
lings were from the Saddle Mountain stand — a population 
from high elevation. There is also high variation among 
families within stands (table 7). Also, characteristics such 
as red coloration of the succulent stem and the prolifera- 
tion of lateral buds vary with population (fig. 5). So far, 
none of this variation can be associated with geographic 
position or elevation. 

Inbreeding — In many stands, only a few individuals 
remain. This results in much higher opportimity for self- 
fertilization (inbreeding). At what level will inbreeding 
adversely affect growth and survival of the seedlings? 
A study has been initiated to compare growth and other 
traits between a highly decimated population and a stand 
with very low mortality. Trees at the two sites will be in- 
tercrossed, crossed with pollen from the same site, and 
compared v^dth wind-pollinated seed. 



BREEDING PLANS TO DEVELOP 
RESISTANCE 

Our surveys have shown that there are many pheno- 
typically resistant whitebark pine trees. A traditional 
breeding approach could be used to develop a new blister 
rust-resistant variety of whitebark pine from these trees. 
Three options in this approach will be discussed. How- 
ever, because of the high level of natural regeneration of 
whitebark pine in clearcuts and burns adjacent to high- 
mortality stands, another approach appears more attrac- 
tive. This is a more natural method using mass selection 
as the genetic selection system, but within a natural set- 
ting and with natural selection processes. 

The Traditional Approach (fig. 6) — Steps to success 
are: 

1. Locate blister rust-resistant phenotypes of white- 
bark pine. Most stands have a few candidate trees, but 
a high number is required so that we can be assured of 




Figure 4 — Three-year-old container-grown white- 
bark pine. 



122 



Table 7 — Variation among whitebark pine families within the 
Gisborne Stand 



'amily 


Third-year 
neignt 


Red stem 


Prolific 

DUO 


Diister 
rust 




cm 




- - Percent — 












1 


18 


6 


0 


11 


2 


12 


11 


0 


22 


3 


16 


0 


0 


29 


4 


14 


12 


0 


0 


5 


15 


20 


5 


22 


6 


14 


7 


0 


14 


7 


17 


0 


0 


7 


8 


15 


4 


13 


33 


9 


18 


20 


4 


65 


10 


15 


9 


0 


35 


11 


12 


5 


9 


30 


Mean 


15 


9 


3 


24 



finding several genes for resistance. Thus, many stands 
will have to be visited. The breeding population for west- 
em white pine is over 3,000 phenotypically resistant trees 
and is considered a minimum nimiber of parent trees to 
assure maintenance of resistance against several races of 
the rust fungus. 

2. Collect wind-pollinated seed from each tree. In spe- 
cial cases, where there are just one or two trees, artificial 
pollination by other candidate trees may be advisable. 

3. Sow seed, grow seedlings, and inoculate with blister 
rust in the fall of their second growth period. 

4. Data over the next 4 years, after inoculation, will 
reveal the most resistant parents of the phenotypically 
resistant candidate trees, along with their resistant 
progeny. 

Options for developing the new resistant variety are: 

1. Selected resistant parents can be grafted into a seed 
orchard. 

2. Resistant seedlings can be used to establish a seed 
orchard. 

3. Surviving seedlings from rust tests that were not 
used in the seed orchard can be outplanted in a natural 
site. The numbers for this planting can probably be in- 
creased by rooted cuttings of the surviving seedlings. 

Timing: The first five steps cotdd be done in about 10 
years. Time to flowering of grafts could be 10 years, but it 
will probably take from 40 to 50 years for their progeny to 
flower. New technology may shorten this time. 

Problems: 

1. What is the natural level of resistance? 

2. How far can seed be transferred without serious 
maladaptation? 

Natural Selection Stand — In many localities in Idaho 
and Montana whitebark pine has been decimated by blis- 
ter rust, leaving just a few individuals per hectare. Many 
of these remnant stands are adjacent to burns or clear- 
cuts. In at least several dozen cases these openings 



contain high nvunbers of whitebark pine seedlings and 
saplings (table 5). With careful management we may be 
able to use these newly established stands for our piir- 
pose. Only three major steps need to be taken: 

1. Select the areas within the clearcuts and bxarns that 
will be managed for whitebark pine (fig. 7A). 

2. Clean the selected area of competing trees and 
shrubs (fig. 7B). Whitebark pine is a poor competitor. In 
fact it probably can be grown at much lower elevations if 
the site is kept free of competition. 

3. Let nature take its course concerning the selection of 
the most resistant trees (figs. 7C, D). 

Problems: 

1. What is the natural level of resistance? We need to 
know the level of resistance (1, 5, or 10 percent) in stands 
to choose sites that have a good chance for success. If only 
1 percent of the seedlings are resistant when produced by 
a mature stand that has had 90 to 95 percent mortality 
by blister rust, a candidate area woidd have to have an 
unusually high number of seedlings to end up with a rea- 
sonably stocked stand. If 10 to 20 percent are resistant, 
the probability of producing a new stand would be consid- 
erably better. 




Figure 5 — Proliferated lateral buds on whitebark pine. 



123 



Locate Parents 
(Resistant Phenotypes) 

Collect Wind or 
Cross-Pollinated Seed 

Blister Rust Test 
1 0 Years 



Graft Seed Orchard 
Production 
1 0 Years 



Seedling Seed Orchard 
Production 
40 Years 



Outplant Survivors 
From Blister Rust 
Test 



Stand Establishment 
50 Years 



Stand Establishment 
50 Years 



Stand Establishment 
50 Years 



Total Time 
70 Years 



Total Time 
100 Years 



Total Time 
60 Years 



Figure 6 — Options for traditional breeding methods for developing new varieties of blister rust-resistant 
whitebark pine, including completion times. 



2. How much inbreeding is occurring in the decimated 
stands? Severe inbreeding might result in reduced 
growth and survivabihty. 

3. What is the minimimi number of cones £ind seed 
needed before seed caches are effective for regeneration? 
Some sites may be beyond help. 

4. When these sites start producing resistant seed how 
far can the seed be transferred? 

Timing (fig. 8): First-generation blister rust selection 
has already been made on parent trees. Seedling estab- 
lishment has already occurred on dozens of sites. Natursd 
selection by blister rust is occmring. In 40 to 50 years es- 
tablished stands on preferred sites could be producing 
cones. 

However, it is not only the increase of resistance genes 
that is important — although this is a key to the species' 
future — it is important to maintain as much of the natu- 
ral function of whitebark pine forests as possible. In addi- 
tion to the severe genetic impact to whitebark pine itself, 
the critical ecological functions of this species have been 
threatened by loss of integrity of whitebark pine forests. 
Thresholds are a key consideration when attempting to 
restore severely altered ecological systems (Friedel 1991; 
Laycock 1991). The ecological thresholds for whitebark 
pine forests are not well understood and with the loss of 
most of the original structure of these forests much of the 
integrity has been lost. It will take generations to repro- 
duce nearly pure, mature whitebark pine forests. In the 
interim, components of the ecological systems most af- 
fected by the loss of whitebark pine forests may be irre- 
versibly altered. 

Time is of the essence, but all is not lost. Over large ar- 
eas the first generation is mostly gone, but in many areas 



the second generation is there; we must act now to save it. 
In other sireas the first generation is largely intact. In 
these areas immediate emphasis needs to be placed on re- 
taining or restoring the ecological function of whitebark 
pine. Protective measures, such as pruning and Ribes 
population manipidation may be used. 

What we are trying to say in this section is that there is 
much that can and must be done immediately to maintain 
and restore whitebark pine in the ecosystem. 

BLISTER RUST ON WHITEBARK 
PINE, A GENETICS PERSPECTIVE 

The general level of resistance of the original popula- 
tion (before the introduction of blister rust) is less than 
1 percent. Therefore, the genetic consequence of the blis- 
ter rust epidemic for most of the range of whitebark pine 
is to push it through an extremely small bottleneck. In 
some areas where the number of surviving trees has been 
reduced to only a few over several hundred hectares, the 
bottleneck may result in entire populations being lost. 
Figure 1 indicates geographic populations of whitebark 
pine that will have the most immediate difficulty (high in- 
cidence area), those that will have only minor problems 
(moderate incidence area), and those that are not likely to 
have a problem (little or no incidence). Fiulher research 
clarifying levels of natural resistance will be a significant 
aid to predicting the adequacy of natural selection in the 
various areas. On some sites, time alone may return 
whitebark pine to its former levels; on other sites, white- 
bark pine may be lost if not aided by genetic improvement 
and silvicultural programs. 

In areas where there are many surviving trees within a 
very reduced popidation, the concern will be the damage 



124 




Whitebark Pine Competing Species 

Figure 7 — Development of a natural selection stand: A, A clearcut before cleaning containing whitebark 
pine and competing trees; B, clearcut after cleaning; C, clearcut after infection by blister rust and selection 
of most resistant trees; D, the natural selection stand or seed orchard after growth of selected trees. 



Locate Sites 
Remove Competing 
Trees 



Stand Establishment 
50 Years 



Some stands are 
already 20 years old 



Total Time 
50 Years or Less 



Figure 8 — Natural selection method for developing 
a new vahety of blister rust-resistant whitebark pine, 
including completion times. 



done by inbreeding. In most conifers inbreeding is detri- 
men1:al. Inbred trees generally grow slowly, and detrimen- 
tal genes such as for albinism, short leaves, and stunted 
seedlings may often occur. However, whitebark pine is of- 
ten transported by birds to areas where there is no other 
whitebark pine (Linhart and Tomback 1985). Because 
there are only a few individuals in the start-up stand, in- 
breeding would be high. This may have been the most 
common way of past population expansion and, over time, 
lethal and other detrimental genes may have dropped out 
of the population leaving a species that is not adversely 
affected by inbreeding. 

Even vTith all these dire genetic consequences, the ge- 
netic opportimity for the survival of whitebark pine in the 
high-mortality zone appears very high, especially with the 
intelHgent help of humans. There are many phenotypi- 
cally resistant trees. These surely can be used as a base 



125 



for new varieties of whitebark pine resistant to blister 
rust. But most exciting, the high natural regeneration of 
whitebark pine stands in clearcuts and burns — even in ar- 
eas where there appear to be insufficient nimibers of par- 
ent trees — indicates that natural processes are already 
producing a new variety. Further, these trees are grow- 
ing well; there do not appear to be inbreeding problems. 

CONCLUSIONS 

The consequences of the blister rust epidemic have to be 
viewed as a disaster for whitebark pine, and a setback for 
Clark's nutcrackers, grizzly bears, and other components 
of whitebark pine forest ecosystems. Nevertheless, white- 
bark pine appears to be equipped with elements that will 
permit survival. Most important is seed caching by the 
nutcracker even when there is just a small excess. Also, 
phenotypically resistant trees seem to be fairly common. 
Considering the low numbers of noncankered western 
white pine trees together with the much higher suscepti- 
bility of whitebark pine, we wonder that there are any 
whitebark pine trees at all. 

Over large areas, the first generation of whitebark pine 
is nearly gone; however, the second generation is often 
present. In other areas the first generation is still largely 
intact. To save whitebark pine and its ecosystem in the 
most sensitive areas it is essential that we act now. Main- 
taining sufficient ecosystem function for whitebark pine 
while developing greater resistance will be a challenge for 
land managers. 

REFERENCES 

Arno, S. F.; Weaver, T. 1990. Whitebark pine community 
types and their patterns on the landscape. In: Schmidt, 
W. C; McDonald, K J., comps. Proceedings — symposium 
on whitebark pine ecosystems: ecology and manage- 
ment of a high-moimtain resource; 1989 March 29-31; 
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Research Station: 97-105. 

Bedwell, J. L.; Childs, T. W. 1943. Susceptibility of white- 
bark pine to blister rust in the Pacific Northwest. Jour- 
nal of Forestry. 41: 904-912. 

Bingham, R. T. 1972a. Artificial inoculation of large num- 
ber of Pinus monticola seedlings with Cronartium 
ribicola. In: Bingham, R. T.; Hofi", R. J.; McDonald, 
G. I., eds. Biology of rust resistance in forest trees. 
Misc. Publ. 1221. Washington, DC: U.S. Department of 
Agriculture: 357-372. 

Bingham, R. T. 1972b. Taxonomy, crossability, and rela- 
tive blister rust resistance of 5-needled white pines. In: 
Bingham, R. T.; Hoff, R. J.; McDonald, G. I., eds. Biol- 
ogy of rust resistance in forest trees. Misc. Publ. 1221. 
Washington, DC: U.S. Department of Agric;ilture: 
271-280. 

Bingham, R. T. 1983. Blister rust resistant western white 
pine for the Inland Empire: the story of the first 25 
years of the research and development program. Gen. 
Tech. Rep. INT-146. Ogden, UT: U.S. Department of 
Agriculture, Forest Service, Intermovmtain Forest and 
Range Experiment Station. 45 p. 



Childs, T. W.; Bedwell, J. L. 1948. Susceptibility of some 
white pine species to Cronartium ribicola in the Pacific 
Northwest. Journal of Forestry. 46: 595-599. 

Childs, T. W.; Bedwell, J. L.; Englerth, R. H. 1938. Bhster 
rust infection of Pinus albicaulis in the Northwest. 
Plant Disease Reporter. 22: 139-140. 

Delatour, C; Birot, Y. 1982. The international lUFRO ex- 
periment on resistance of white pines to blister rust 
{Cronartium ribicola): the French trial. In: Heybroek, 
H. M.; Stephan, B. R.; Weissenberg, K., eds. Resistance 
to diseases and pests in forest trees. Wageningen: 
Pudoc: 412-414. 

Friedel, M. H. 1991. Range condition and the concept of 
thresholds: a viewpoint. Journal of Range Management. 
44: 422-426. 

Hoff, R. J.; Bingham, R. T.; McDonald, G. I. 1980. Relative 
blister rust resistance of white pines. European Journal 
of Forest Pathology. 10: 307-316. 

Laycock, W. A. 1991. Stable states and thresholds of 
range condition on North American rangelands: a view- 
point. Journal of Range Management. 44: 427-433. 

Linhart, Y. B.; Tomback, D. F. 1985. Seed dispersal by 
nutcrackers causes multi-trunk growth form in pines. 
Oecologia. 67: 107-110. 

McCaughey, W. W.; Schmidt, W. C. 1990. Autecology of 
whitebark pine (Pinus albicaulis Engelm.). In: Schmidt, 
W. C; McDonald, K J., comps. Proceedings — symposium 
on whitebark pine ecosystems: ecology and manage- 
ment of a high-mountain resource; 1989 March 29-31; 
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Research Station: 85-96. 

McDonald, G. I.; Hoff, R. J. 1970. Resistance to Cro- 
nartium ribicola in Pinus monticola: early shedding of 
infected needles. Res. Note INT-124. Ogden, UT: U.S. 
Department of Agriculture, Forest Service, Intermoun- 
tain Forest and Range Experiment Station. 8 p. 

Pfister, R. D.; Kovalchick, R. L.; Arno, S.; Presby, R. 1977. 
Forest habitat types of Montana. Gen. Tech. Rep. 
INT-34. Ogden, UT: U.S. Department of Agriculture, 
Forest Service, Intermoimtain Forest and Range 
Experiment Station. 174 p. 

Spaulding, P. 1911. The blister rust of white pine. Bull. 
206. Washington, DC: U.S. Department of Agriculture, 
Bureau of Plant Industry. 88 p. 

Steele, R.; Cooper, S. V.; Ondov, D. M.; Roberts, D. W.; 
Pfister, R. D. 1983. Forest habitat t3rpes of eastern 
Idaho-western Wyoming. Gen. Tech. Rep. INT-144. 
Ogden, UT: U.S. Department of Agriculture, Forest Ser- 
vice, Intermountain Forest and Range Experiment Sta- 
tion. 122 p. 

Steele, R.; Pfister, R. D.; Ryker, R. A.; Kittams, J. A. 1981. 
Forest habitat types of central Idaho. Gen. Tech. Rep. 
INT-114. Ogden, UT: U.S. Department of Agriculture, 
Forest Service, Intermountain Forest and Range Ex- 
periment Station. 138 p. 

Stephan, B. R. 1985. Resistance of five-needle pines to 
blister rust. Allgemeine Forstzeitschrift. No. 28: 
695-697. 

Yokota, S. 1983. Resistance of improved Pinas monticola 
and some other white pines to the blister rust fungus, 
Cronartium ribicola, of Hokkaido, Japan. European 
Journal of Forest Pathology. 13: 389-402. 



126 



PERFORMANCE OF PINUS CEMBRA, 
P. PEUCE, AND P. STROBIFORMIS 
WITHIN AIR-POLLUTED AREAS 

Karel Kahak 



Abstract — ^An inventory of the surviving species within the 
heavily air-polluted area of the Ore Mountains showed a number 
of species of the genus Pinus without symptoms of injury. Three 
species of the stone pines are described to demonstrate the im- 
portance of the geological past of the species. It relates to the 
impact of the catastrophes and volcanic activities and sudden 
changes of climate for the species' preadaptation to the toxic 
impact of a changing environment. The causes of the resistance 
seemed to be connected with the long-term migration {Pinus 
cembra, P. peace) and hybrid origin (P. peace, P. strobiformis). 
The importance of the life history of the species is unquestioned. 



During the years 1956 until 1966 the author (Kahak 
1971, 1988, 1991) estabhshed an international collection 
of species of gentis Pinus in Plzeh, West Czechslovakia, 
in the forest locality of Sofronka (latitude 50° N.). 

Various species of pines are represented in Sofronka ei- 
ther by patterns of populations that come from different 
localities within every species range or from groves of in- 
dividual trees from species that originated from locations 
that were not sufficiently preadapted for the local condi- 
tions of the new environment (Kahak 1988). A research 
center was established at the Arboretxim Sofronka in 1963 
to investigate evolution of the species included within the 
collection. The entire range of provenance experiments 
with some of the species was analyzed with respect to the 
evolution of their distribution area during postglacial 
time. 

The air pollution caused by extensive industrial emis- 
sions resulted in a mass extinction of the Norway spruce 
forests in the Ore Mountains in the vicinity of Plzeh. In 
the first stage of the extinction, the dying spruce stands 
were cleared, but small groves of individual trees, pre- 
dominantly the local and exotic species of pines, survived. 
Therefore, the staff of the research center at Sofronka was 
invited to join the search for an ecologically harmless 
method of reconstructing the extinct forests. 

Some species of white pines were discovered whose 
needles appeared to be unaffected by emissions. These 
species of stone pines were: Swiss stone pine {Pinus 
cembra L.), Macedonian pine (Pinus peuce Griseb.), and 
southwestern white pine {Pinus strobiformis Engelm.). 
Their unique viability had helped them to survive under 
the highly unfavorable conditions. 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Karel Ka&fik, National Park of Simiava, Vimperk, Czechoslovakia. 



SWISS STONE PINE 

1. Occurs as individual trees in the entfre emission 
area between the town of Hora Sv. Sebestiana and Mount 
Klmovec (1,210 m above sea level [a.s.l.]). Some trees, 
otherwise injured by snow and icing, show no injury from 
emissions. Many of them bear germinant seeds (Mottl 
and Prudic 1982). 

2. One grove of fertile trees 60 to 70 years of age occurs 
in the cemetery arboretiun in the town of Vejprty. Their 
dimensions are 12 cm diameter at 1.3 m and 9-11 m 
height. They bear germinant seeds. 

MACEDONIAN PINE 

On the Saxon side of the mountains: 

1. Forest district Steinbach, Department 157a2, eleva- 
tion 800 m a.s.l. A group of 17 trees, age of 55 years, 
fertile, with natural regeneration. Dimensions: Diameter 
26-35 cm at 1.3 m and 15 m height. 

2. Forest district Schmalzgrube, Department 291a3, 
elevation 900 m a.s.l. Twelve trees of the same prove- 
nance and age as the latter growing on a peat bog! The 
trees are healthy and fertile. Dimensions: Diameter 
15-40 cm at 1.3 m and 11-14 m height. Tree tnmks have 
been injured by the red deer and humans (chmbing 
sptars). 

3. Forest district Grumbach, Department 331a4, ele- 
vation 825 m a.s.l. Stand area is 0.75 ha and trees are 
50 years old (1982) and grow in a mixture with Pinus 
contorta ssp. latifolia and the movmtain variant of Pinus 
sylvestris L. Average height of the stand is 18 m, diam- 
eter is 21-35 cm at 1.3 m. Macedonian pine is the most 
vital of the named species. 

On the Czech side of the mountains: 

1. Locality Bludna, Forest Management Horm' Blatna, 
elevation 1,000 m a.s.l. Northern slope stand of approxi- 
mately 200 trees, age probably 70 years. Origin tmknown, 
but probably from activities of the Forest Research Insti- 
tute Mariabnmn just before World War I. Many trees were 
apparently broken at the height of 5-7 m, but replacement 
crowns are now reaching heights of 15 m. Trees are rela- 
tively fertile with natural regeneration present. 

SOUTHWESTERN WHITE PINE 

1. Locality: Cemetery arboretum in the town of Vejprty 
with a heavy emission impact. One tree only. Diameter 
36 cm at 1.3 m, height 14 m, and age 50-60 years. Tree is 
fertile with germinant seeds in the cones Ijdng on the 
moist grass! 



127 



PROBABLE CAUSES OF RESISTANCE 
Autochtonous Species 

Ten years of observations and experimental activi- 
ties in the 7-year period 1982-89 give us the following 
indications: 

Variability — ^Among the local forest tree species, there 
are different degrees of resistance between individual 
trees within the stands. Both autochtonous and alloch- 
tonous pine species had been gradually reduced, but there 
are some examples without any symptoms of injury. Fir 
disappeared on the Czech side of the mountain range 
first, but drought rather than emissions may have been 
the cause. Norway spruce seems to be extinct. 

Catastrophic Selection — Norway spruce mostly dis- 
appeared in a relatively short time — about 6 years. How- 
ever, after its massive extinction, individually resistant 
trees of this species were found within the extinct forests. 
This appearance is commonly known under the term 
"population bottleneck" (Dobzhanski and others 1976) or 
"catastrophic selection" (Lewis 1962; Raven 1964) or "une 
evolution acceleree (Bouvarel 1960). 

Founder Effect — ^According to Mayr's "founder-effect" 
(Mayr 1942, 1965, 1979) after the strong reduction of 
Norway spruce the reconstruction of its genetic architec- 
ture within the small number of survivors (founder popu- 
lation) might follow (Kahak 1988). This might give rise 
to a new species population endemic to the environmental 
conditions that caused the latter extinction. 

Allochtonous Species 

Hybridization — The Canadian species of spruce show 
a high degree of emission resistance compared to the local 
Norway spruce. Their life history, with their intricate 
genetic relations, may help explain this orientation 
(Heimburger and others 1983; Morgenstern and Farrar 
1964). Morphological and genetic evidence indicates that 
black spruce is a relatively young species. It is thought to 
be of an American origin resulting from the hybridization 
of a proto- white spruce from eastern Asia migrating east- 
ward with a proto-red spruce of North American origin 
migrating westward. Differences in morphological char- 
acteristics between black spruce and red spruce in the di- 
rection of white spruce would indicate such hybrid origin 
(Heimburger 1983). 

Directional Selection 

The most relevant information came from all the for- 
eign species formerly introduced to this mountain range. 
An excellent example is the life history of lodgepole pine 
as described by Hansen (1942, 1943). This species came 
from the Cascade Range territories of North America 
where it had been influenced by repeated volcanic activ- 
ity. Consequently, many generations of lodgepole pine 
were subjected to directional selection through the toxic 
air and soil, and lodgepole pine adapted to become the 
founding pioneer stage of the ecosystems' evolution. 



Migration Preadaptation 

The other very important evolutionary factor is migra- 
tion. It acts on the development of ecological potentisds 
as a consequence of the tolerance to the extreme envi- 
ronmental conditions (toxic sites), the environmental 
changes, and all the preadaptations that could be ac- 
quired only during a long migration route such as those 
in the Tertiary — one from eastern Asia to southern 
Europe (Mirov 1967) and the other from eastern Asia 
to Australia (Mayr 1965). 

HISTORY OF SPECIES 

Mirov (1967) located the center where the genus Pinus 
originated in Eastern Asia, where catastrophes like vol- 
canic activity and changes in climate gave all the species 
of pines a specific starting point to their principal 
adaptation. 

"Pines however possessed two characteristics since the 
Mezozoic origin of the genus: their xeromorphy and their 
ability to endure direct sunlight. These characteristics 
may be considered as generally the same throughout the 
whole paleobotanical history." (Mirov 1967). 

Pinus cembra 

"During the Ice Age, the high mountain pines of the 
Alps either perished or were preserved in the lower areas 
not covered with ice. Pinus cembra found occasionally in 
northern Italy is the relic preserved in one of such refugia 
(Emberger 1944). With subsequent warming up, it mi- 
grated again to the high elevations of the central Alps. 
There is, of course, a possibility of the existence of non- 
glaciated refugia even in the central parts of the Alps. 
Pinus cembra might have survived in some of these few 
areas that remained free from the ice." (Mirov 1967). 

The indications are that P. cembra and many other spe- 
cies of the eastern Asia flora migrated to the Alps during 
the Tertiary. During the last glaciation they occurred in 
an almost continuous belt of higher elevations. Conse- 
quently, because of the enlarged area of its distribution, 
P. cembra developed a well-supported degree of genetic 
variability. This species of stone pine survived the last 
glaciation, and within its refugias there was probably an 
increasing frequency of autogamy that must be taken into 
account (Bannister 1965). After the retreat of the glacier, 
the migrating progenies probably intercrossed at their 
contact zone and thus acquired their increasing variabil- 
ity as well. Together with the very short period of the 
last glaciation, the imdiminished variability still main- 
tains itself today. Pinus peuce apparently has a similar 
pattern of the life history. 

Pinus peuce 

According to Mirov, "There are indications that Medi- 
terranean pines migrated from eastern Asia along the 
mountain ranges, that once extended north of and parallel 
to Himalaya." (Mirov 1967). 

"Endemic Pinus peuce is very local in Bulgaria, Albania 
and Yugoslavia. It is a haploxylon pine closely related to 



128 



Pinus griffithii of the Himalaya. Before the great migra- 
tion, Pinus pence has been crossed with other haploxylon 
pines. Judging by the chemical composition of its tiirpen- 
tine (Illoff and Mirov 1956) it has a close affinity to east- 
em Asiatic and western American pines" (Mirov 1967). 

The hybridization should be a source of variation for ad- 
aptation to new environments (Lewontin and Birch 1966). 

The whole Mediterranean region experienced many tec- 
tonic changes throughout the Tertiary. In the Quarter- 
nary there occurred intensive faulting and sinking of 
considerable areas. This resiilted in the formation of sep- 
arate moimtain ranges and many islands. As a conse- 
quence, the present distribution of pines in the Mediter- 
ranean region is sporadic and irregular (Mirov 1967). 

Pinus strohiformis 

Southwestern white pine appear to have morphological 
indications of a hybrid origin. Critchfield and Little 
(1966) said: "This species forms a link both geographi- 
cally and morphologically, between the neighbors Pinus 
flexilis and Pinus ayacahuite. Southwestern pine is ap- 
parently separated from its southern relative Pinus aya- 
cahuite by a 100 mile gap that does not seem to corre- 
spond to any comparable discontinuity in suitable 
habitats." 

We have popvilation samples of Pinus strohiformis in 
our arboretum that originated from Greenlee, AZ, at lati- 
tude 33°39' N., longitude 109°16' W., elevation 2,740 m 
a.s.l. This stand is now 30 years of age. Its fertility began 
at the age of 22 years. Its cones are surprisingly variable 
in size, from the smaller cones similar to Pinus flexilis to 
very big cones with apophyses, similar to Pinus aya- 
cahuite var. brachyptera. Such a variation supports the 
hypothesis about the hybrid origin of this species. 

TESTS OF SPECIES POTENTIALS 

The response of a species to the impact of an environ- 
mental stress allows us to recognize intrinsic characteris- 
tics and appearances that we do not usually encounter un- 
der normal circumstances. Therefore, we are focused on 
testing the collection of pine species at the arboretum 
within the zone where there is a heavy impact of both 
emissions and severe mountain climates. 

CONCLUSIONS 

The threatened trees of some forest species imder heavy 
emissions often show a surprising ability to endure the se- 
lection pressure of the toxic environment. Evidently this 
depends on the life history of the species in question. 
Some conditions under which we feel the tested species 
have a better predisposition to siirvive are: 

1. The species is of hybrid origin. 

2. The species has imdergone a long-term interconti- 
nental migration. 

3. The species has been subjected to repeated strong 
directional or even catastrophic selection under the im- 
pact of volcanic activity. Volcanic areas are noted for 
their toxic air and soil similar in some ways to industrial 
emission areas. 



REFERENCES 

Bannister, M. H. 1965. Variation in the breeding system 
of Pinus radiata. In: Baker, H. G.; Stebbins, G. L., eds. 
The genetics of colonizing species. In: Proc. First Int. 
Un. Biol. Sci. Symp. Gen. Biol. Davis; 1964 February. 
New York and London: Academic Press: 353-374. 

Bigelow, R. S. 1965. Hybrid zones and reproductive isola- 
tion. Evolution. 19: 449-508. 

Bouvarel, P. 1960. Note sur la resistance au froid de 
quelques provenances de pin maitime. Revue forestiere 
francaise. 7: 495-508. 

Critchfield, W. B.; Little, E. L., Jr. 1966. Geographic dis- 
tribution of the pines of the world. Misc. Publ. 991. 
Washington, DC: U.S. Department of Agriculture, 
Forest Service. 97 p. 

Diels. 1910. ex Mirov (1967). 

Dobzhanski, Th.; Ayala, F.; Stebbins, L. G.; Valentine, 

L. W. 1976. Evolution. San Francisco: Freeman. 
Emberger. 1944. ex Mirov (1967). 

Hansen, H. 1942. The influence of volcanic eruptions upon 

post-Pleistocene forest succession in central Oregon. 

American Journal of Botany. 29: 214-219. 
Hansen, H. 1943. Paleoecology of the sand dune bogs on 

the southern Oregon coast. American Joiirnal of Botany. 

30: 335-340. 

Heimburger, C. C. 1983. The evolution of Black spruce. 

Proc. XDC. Meet. CTIA, Part 2: 163-166. Toronto. 
Illoff, P. M., Jr.; Mirov, N. T. 1956. Composition of gimi 

turpentines of pines. XXV. A report on two white pines: 

Pinus koraiensis from Korea and Pinus peuce from 

Macedonia. Amer. Pharm. Ass. J. Sci. Ed. 45: 77-81. 
Kahak , K. 1971. Arboretum Sofronka. Introduction 

results in 1958-1968. Praha. 38 p. illus. 
Kahak, K 1988. Contribution to maintaining continuity of 

the Norway spruce in the Ore Moimtains. Folia Mus. 

Rer. Natur. Bohem. Occid. Plzai, Botanika 27. 58 p. 
Kahak , K 1991. Colonizing ability of some introduced 

species of pines. FoKa dendrologica. 18: 187-200. Veda, 

Bratislava. 

Lewis, H. 1962. Catastrophic selection as a factor in spe- 
ciation. Evolution. 16: 257-271. 

Lewontin, R. C; Birch, L. C. 1966. Hybridization as a 
source of variation for adaptation to new environments. 
Evolution. 20: 315-336. 

Mayr, E. 1942. Systematics and the origin of species. New 
York: Columbia University Press. 

Mayr, E. 1965. Summary. In: Baker, H. G.; Stebbins, G. 
L., eds. The genetics of colonizing species. In: Proc. First 
Int. Un. Biol. Sci. Symp. Gen. Biol. Davis; 1964 Febru- 
ary. New York and London: Academic Press: 553-562. 

Mayr, E. 1979. Evolution xmd Vielfalt des Lebens. 
Springer-Verlag. 

Mirov, N. T. 1967. The Genus Pinus. New York: Ronald Press. 

Morgenstem, E. K.; Farrar, J. L. 1964. Introgressive 
hybridization in red spruce and black spruce. Tech. 
Rep. 4. Faculty of Forestry, Univ. of Toronto. 

Mottl, J., Prudic, Z. 1982. Analyza riistu limby na 
Klmovci — V Krusnych horach a jejf roubavanci. 
Tschechisch. Zpravy les. vyzkumu 27(4): 9-13. 

Parsons, P. A. 1963. Migration as a factor in natural selec- 
tion. Genetica. 33: 184-206. 

Raven, P. 1964. Catastrophic selection and edaphic 
endism. Evolution. 17: 336-338. 



129 



COMPETITION AND CROWN 
CHARACTERISTICS OF WHITEBARK 
PINE FOLLOWING LOGGING IN 
MONTANA, U.S.A. 

Todd Kipfer 
Katherine Hansen 
Ward McCaughey 



Abstract — This study was designed to evaluate crown develop- 
ment of whitebark pine (Pinus albicaulis) subjected to intertree 
competition in young mountainous stands in the Yellowstone 
area. A distance-dependent competition index was used to assess 
intertree competitive effects on crown growth. Significant in- 
verse correlations were found between the competition index and 
height, crown diameter, crown diameter/height, and crown vol- 
ume; however, only 20 percent of the variation was explained by 
measured variables. Competition thresholds, indicating when 
crown growth becomes limited, were estimated. In future stud- 
ies, crown vigor or density measures could perhaps yield a better 
competition indicator. 



This study addresses the response of regenerating 
whitebark pine {Pinus albicaulis) to competition following 
logging. Human activities, such as logging, in the movm- 
tainous environments of western North America are cre- 
ating a variety of impacts at both the individual scale and 
at the ecosystem scale. Concern about regeneration 
crown growth and future cone production in logged sites 
prompted this study. Cone production is a product of 
mean shoot production multiplied by the number of fertile 
shoots (Weaver and Forcella 1986). Effects of competition 
were hypothesized to affect crown development of white- 
bark pine. Intertree competition was measured on indi- 
vidual, regenerating whitebark pine in logged sites in 
southwestern Montana near Yellowstone National Park. 
This study is part of a larger research project conducted 
by Kipfer (1992). 

Whitebark pine are found on sites with a wide range 
of geologic, geomorphic, and ecologic conditions (Hansen- 
Bristow and others 1990). Their growth characteristics 
depend on numerous factors, including competition, that 
vary spatially and temporally. Quantifying the influence 
of competing vegetation on conifer grov^i^h may help to de- 
termine silvicultural prescriptions such as thinning inten- 
sity, optimize management decisions, and provide a basis 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Todd Kipfer is a Doctoral Student, Department of Geography, Arizona 
State University, Tempe, AZ 85281; Katherine Hansen is an Associate 
Professor of Geography, Department of Earth Sciences, Montana State 
University, Bozeman, MT 59717; Ward McCaughey is a Research For- 
ester, USDA Forest Service, Intermountain Research Station, Forestry 
Sciences Laboratory, Bozeman, MT 59717. 



for growth models of young forest stands (Wagner and 
Radosevich 1991b). For example, one management option 
for whitebark pine is to increase or sustain seed produc- 
tion for regeneration and for a food source for the Clark's 
nutcracker (Nucifraga columbiana), the red squirrel 
{Tamiasciurus hudsonicus), the grizzly bear (Ursus 
arctos horribilis), and a host of other birds and mammals 
(Craighead and others 1982; Kendall 1983; Kendall and 
Amo 1990; McCaughey and Schmidt 1990). Seed produc- 
tion may increase with v^dder, more diflFuse crowns (Eggers 
1986; Spurr and Barnes 1980), and, therefore, silvicultural 
prescriptions that increase crown diffiisivity are desirable. 
The effects of silvicultural management options for main- 
taining or enhancing whitebark pine ecosystems have had 
limited study (Eggers 1990; Schmidt and McCaughey 
1990). 

PREVIOUS STUDIES 

Whitebark pine is considered relatively intolerant of com- 
petition and shade (Amo and Weaver 1990), although the 
trees may be more shade tolerant at an early stage and 
less tolerant at later developmental stages (McCaughey 
and Schmidt 1990). Although initial observations by 
Eggers (1990) indicate that suppressed whitebark pine 
seedlings and saplings respond little to the removal of for- 
est competition, the effect of competition on postlogging 
regeneration needs further study. Weaver and others 
(1990) assessed competition in terms of stand develop- 
ment and hypothesized that although the density of 
whitebark pine seedlings remained approximately con- 
stant across stands of different ages, growth from seedling 
to sapling size could not be supported in stands older than 
100 years. They suggested that clearings needed to be 
larger than 10 m^ for seedlings to grow to sapling size. 

The effect of competition on the growth of individual 
trees has been studied for many economically important 
trees in plantations or homogeneous stands (Daniels and 
others 1986; Tome and Burkhart 1989; Wagner and 
Radosevich 1991a,b). These studies have primarily used 
individual tree grovvi:h models based on competition indi- 
ces and have shown an inverse relationship between the 
amount of competition from neighbors and individual tree 
growth performance (Daniels 1976; Daniels and others 
1986; Lorimer 1983; Tome and Burkhart 1989). This 
study focused on crown development of whitebark pine 
as related to a distance-dependent competition index. 
Managers could use this index to determine the optimum 



130 



spacing or number and spatial arrangement of competi- 
tors for maximimi crown development of whitebark pine, 

METHODS 

Three logged stands were sampled in the Gallatin Na- 
tional Forest in southwestern Montana near Yellowstone 
National Park. The stands ranged in elevation from an 
average of 2,260 m at Moose Creek to 2,475 m and 2,535 m 
at Teepee Creek 1 and 2, respectively. The three stands 
had been clearcut between 1968 and 1972. Data collec- 
tion focused upon intra- and intertree characteristics for 
whitebark pine and neighboring trees. One hundred 
whitebark pine individuals (more than 0.5 m tall) were 
identified within each of the three stands by a systematic- 
random sampling procedure along transects. Three 
different-sized circular plots were nested around each in- 
dividual whitebark pine. Within each circular plot, trees 
of specific sizes were identified as competitors using the 
following criteria: within the 3-m-radius plot, all tree 
seedlings and saplings were identified as competitors; 
within the 6-m-radius plot all trees with a diameter at 
breast height (d.b.h.) greater than or equal to 4 cm and 
less than 10 cm were identified as competitors; and within 
the 9-m-radius plot all trees with d.b.h. greater than or 
equal to 10 cm were identified as competitors. This 
method approximated a fixed angle gauge sweep (such as 
that used by Tome and Burkhart 1989) and requires that 
trees farther away from the individual tree be larger to be 
considered competitors. 

Five attributes were measured for each whitebark pine: 
(1) age, (2) height, (3) diameter at breast height (d.b.h.), 
(4) average live crown diameter, and (5) height of the 
maximum crown diameter above the ground surface. 
Crown and height variables were measured to the nearest 
0.1 m and d.b.h. to the nearest 0.1 cm. 

Total sample size was reduced from 300 to 220 after 
tree ages showed that some selected whitebark pine indi- 
cated some had established prior to logging (advance re- 
generation). It is difficult to evaluate intertree competi- 
tion prior to logging, and these were therefore eliminated 
from further study. A ratio of crown diameter to tree 
height iCD/H) was computed to reflect the degree of hori- 
zontal crown growth. Crown volume was computed using 
the cone volume equation: (Volume = 1/3* PI * R'^ * H). 
Measurements on competing trees within the 9-m-radius 
plot were tree species, total tree height, diameter at 
breast height, and distance (measured to nearest 0.1 m) 
to the subject whitebark pine. 

A distance-weighted size-ratio index or distance- 
diameter index (Alemdag 1978; Daniels 1976; Hamilton 
1969; Hegyi 1974; Tome and Burkhart 1989) was selected 
to develop a competition index, from which intertree com- 
petition could be quantified. The distance-diameter index 
is a distance-dependent index that svuns a size ratio of 
diameters (d.b.h. or basal diameter) of competing trees to 
the diameter of a subject tree. The general theory is that 
larger competing trees, relative to the individual tree, 
contribute more competition than do smaller trees. The 
index also assumes that competitive influence decreases 
with increasing distance between the individual tree and 
the competitor. 



The distance-diameter index was modified for white- 
bark pine because the trees often had multiple stems or 
did not reach breast height. Height was substituted for 
diameter and distance weighting functions were applied 
to develop the modified competition index: 

N 

C.I.= Z (if./i/.)*function(Z)/ST..) 
i=l ' ' 

where C.I. is the competition index; if. is the tree height 
of the subject tree; H. is the tree height of the competing 
tree; and DIST.^ is a distance decay function. Three dis- 
tance decay functions were used: {1/DIST , l/DIST^ , and 
e-°'s^y). 

Regression analysis was used to evaluate the relation- 
ship between competition and crown characteristics of the 
whitebark pine. The three distance-decay functions were 
used as dependent variables and tree height and three 
crown measures (crown diameter, crown diameter to tree 
height ratio, and crown volimie) as independent variables. 

RESULTS 

Competition index values varied according to the form 
of the distance-decay function used for whitebark pine. 
Index values calculated using 1/distance and 1/distance^ 
were similar with mean values ranging from 13.12 to 
17.12. Index values using the e~^'^^'^^ decay function 
ranged from 3.11 to 4.22. 

Tree heights of whitebark pine were significantly corre- 
lated (P = 0.007) with the competition index; however, 
values for the three distance-decay functions (0.033 for 
1/distance, 0.027 for l/distance^, and 0.032 for e"^''*^^^) 
indicate that the linearity of that relationship is poor as 
shown for 1/distance in figure 1. The intercept value in 
the regression equation (y = a + bx) was 2.047, which may 
represent maximimi tree heights for whitebark pine in 
the age classes represented by this study, when growing 
in the absence of competition. 

Crown diameters of whitebark pine were significantly 
correlated {P = 0.000) with the competition index, shown 
for 1/distance in figure 1, with R^ values varying depend- 
ing on the form of the distance-decay function (0.131 for 
1/distance, 0.096 for l/distance^, and 0.128 for e"^*^^^^). 
The intercept value (0.820) of the regression equation 
poorly represents the maximum possible crown diameter, 
since crown diameter values above 1.0 m were common. 

The crown diameter to tree height ratio and crown vol- 
ume were significantly correlated to competition index as 
shown for the 1/distance decay function in figure 1. Re- 
gression R^ values for crown diameter to tree height ratio 
varied depending on the form of the distance-decay func- 
tion in the competition index equation (0.122 for 1/distance, 
0.079 for l/distance^, and 0.120 for e-^'^'=^). Similar regres- 
sion i?^ values were obtained in correlating competition 
index with crown volume using the three forms of the 
distance-decay function (0.092 for 1/distance, 0.058 for 
l/distance^ and 0.091 for e-^^^-^^). 

The 1/distance decay function in the competition index 
model yielded slightly higher R ^ values for tree height, 
crown diameter, crown diameter to tree height ratio, and 
crown voliraie of whitebark pine. The 1/distance decay 



131 



S.0- 
4.5' 
4.0 

3.5 ■ 

ao 
^6 

2.0 

1J 

1.0 

as 
ao 



Tro halght » 2.047 - 0.0133'C.I. 
R-squarod > 0.033 
p = 0.007 n > 220 



• • • • 

£ li *•* t .y . r 



* 



— IX)- 



0.0 



Crown diameter > 0.820 - 0.0117*C.I. 
R-tquarad = 0.131 
p - 0.000 n - 220 



• a • a 

• aa « 

a a 

•aaa a a a a 



aaa • a • a a 
■ a aa a a 
a a a 

a a a aa 
•a a «a a 



a aa •aaa 

•aa a aa a« 

a a a a 

aaa a aa • 

a aa>* a 

a a aa • a 

a aaa 



a a 
a a 



a a a 
a a 



Competition Index (I/distance) 



Competition index (l/dietance) 



a 

a 


Crown diameter/Tree height « 0.428 - 0.00481 'CI. 
R-squarad ' 0.122 
p - 0.000 n 3 220 


a 

Sa • • a 
a'« a-a a 
J a a • •• 


a 


aaa a • 
a^ ^'V.-.a ;C 


a • 
• a • • • a 

a a a 


• ^"'^ 

a 

aa 


• 



10 20 30 40 60 

Competition index (I/distance) 



60 



CO 

< 



Crown vokima <s 0.300 - 0.00056*0.1. 
R-aquarad • 0.002 
p « 0.000 n = 220 



# • 



• t • 



•••'a *. 



10 20 30 40 SO 

Competition index (1/distance) 



Figure 1 — Bivariate scatterplots of total tree heiglit, crown diameter, crown diameter/lielght, and crown 
volume of whiitebarl^ pine plotted against competition index using a distance-decay function of 1/distance. 
Regression values shown are model and model coefficients, fl-square, p-level, and sample size (n). (Data 
from all tiiree study sites — iy/loose Creek, Teepee Creek 1 , and Teepee Creek 2.) 



function was used for development of a potential competi- 
tion threshold because it was consistently a stronger pre- 
dictor of the four growth variables. 

Scatterplots were evaluated from each study site for 
competition index values above which there was visually 
less variability in the data. No identifiable threshold was 
observed for whitebark pine in the Moose Creek stand, 
and therefore, in order to increase the sample size, the 
two Teepee Creek stands were combined. Potential com- 
petition thresholds were estimated for total tree height, 
crown diameter, crown diameter to tree height ratio, and 
crown voltmie for sampled whitebark pine in the two Tee- 
pee Creek stands (fig. 2). Variability of data below poten- 
tial competition thresholds indicates that tree growth fac- 
tors may not be limiting. The majority of whitebark had 
competition index values below the estimated threshold 
values. 

Regression analysis of the competition index against 
the four whitebark pine measures using only those trees 
greater than the threshold values jdelded poor results for 



tree height (R^ = 0.128,p = 0.062, n = 27), crown diameter 
(R^ = 0.146, p = 0.106, n = 18), and crown volume (R^ = 
0.091, p = 0.120, n = 27). A statistically significant corre- 
lation, however, was found between the competition index 
and the crown diameter to tree height ratio (CD/H = 
0.846 - 0.0142*C.I.: R^ = 0.230, p = 0.000, n = 18). 

DISCUSSION 

The specific competition index values developed by this 
study explain approximately 20 percent of the variation in 
competitive pressure exerted on whitebark pine from the 
numbers, sizes, and spatial patterns of neighboring trees. 
Assessment of competition on crown development of 
whitebark pine is difficult when data come from present 
spatial relationships and past growing conditions are un- 
known. Competition was a significant factor influencing 
crown characteristics of whitebark pine in this study. It 
is important to note that these results only show that 
competition is a significant factor influencing whitebark 



132 



5.0 
4.5 
4.0 

3.5 

ao 

2.5 
2.0 
1.5 
1.0 
OJ 

ao 



• 




• 

• 

• 


< Potential threshold 


• 

;i.r J» • t, - 

• • • 

• • • 


•• • 

* * • 

• 


••• • •• • 


• • • 

• • 







20 30 40 SO 

Competition index (1/distance) 



E 



OX 



( • •••• 



• • • • • I 



Potential threshold 



10 20 30 40 50 

Competition Index (1/dlstance) 







Potential threshold 


• 

• 


< 


• 

\, • • ' 

• • • • • • 












• • 

m • 
• 


• 

• 

• 

• • • 
m 


• • 







20 30 40 60 

Compstltlon Index (1/dlstance) 



n 
< 

o 

E 

o 
> 



• • •/ 



Potential threshold 



10 20 30 40 £ 

Competition index (I/distance) 



Figure 2 — Potential competition thresholds for bivariate scatterplots of total tree height, crown diameter, 
crown diameter/height, and crown volume of whitebark pine plotted against competition index using a 
distance-decay function of 1/distance. (Data from Teepee Creek 1 and Teepee Creek 2 only.) 



pine. The mechanisms of this competitive influence are 
unknown. 

The distance-decay function 1/distance in the competi- 
tion index model created an expected inverse relationship 
of tree growth measure to index. As distance from the 
subject tree increases effects from competition trees de- 
crease. Tree heights and crown volumes had the poorest 
correlations with the competition index. Poor correlations 
with crown volume may be due to sampHng procedures. 
In future studies, crown vigor and crown biomass meas- 
urements (stem spacing) might be used to weight crown 
volume measures, perhaps yielding a better indication of 
the effect of competition on growth. 

Crown diameters and crown diameter to tree height ra- 
tios had good correlations with the competition index. 
Still, only 12 to 13 percent of the variation in these meas- 
ures was explained by the competition index. The addi- 
tion of age and an indicator variable for stand improved 
this relationship for the crown diameter measure, increas- 
ing the R 2 value to 0.250. It did not, however, yield a sig- 
nificant correlation for the crown diameter to tree height 



ratio. The crown diameter to tree height ratio may repre- 
sent a characteristic crown shape for young whitebark 
pine. 

Poor correlations between whitebark pine crown charac- 
teristics and competition index may be due to factors in- 
fluencing individual tree size (local density, plant geno- 
type, seed size, emergence time, microhabitat variations, 
and unknown historical growing conditions). A competi- 
tion index coxdd incorporate a variety of growing condi- 
tions over time and the use of incremental growth charac- 
teristics of annual growth could provide some improvement 
without substantial measurement efforts. 

Potential competition thresholds estimated for the two 
Teepee Creek sites may be artifacts due to relatively few 
trees with high competition values, or they may be actual 
values of competitive pressure indicating when growth 
becomes limited by competition. Multiple-aged stands 
with varying tree densities should be sampled in future 
studies. To compare stands of different ages, the age- 
independent competition index methods of Lorimer (1983) 
may provide a strong starting point. Although regression 
analysis can comparatively evaluate the influences of 



133 



competitive pressure on whitebark pine characteristics, 
actual relationships are more complex than the linear re- 
gression models might suggest. Nonlinear regression 
methods should be explored for evaluating competition 
influences. 

Timber harvests provide an important but declining 
part of our regional economy (Powers 1991), and economic 
considerations are an important component of forest man- 
agement policies. Timber harvest can cause accelerated 
slope failures, erosion, and stream sedimentation from 
roads (Marston and Anderson 1991) when improperly 
done. Road building in association with logging provides 
access to whitebark pine regeneration sites; however, it is 
not believed that logging will be a major means for regen- 
erating whitebark pine (Arno 1986). 

It is imperative to find viable management options for 
increasing crown growth and the potential for increased 
cone production for whitebark pine because of its impor- 
tance to the endangered grizzly and many other wildlife 
species in the Yellowstone ecosystem. The importance of 
quantifying how competing trees influence crown charac- 
teristics of whitebark pine in logged stands comes from a 
need to increase our knowledge of how competition affects 
crown development of whitebark pine. 

ACKNOWLEDGMENT 

This research was supported in part by funds provided 
by the Intermoxmtain Research Station, Forest Service, 
U.S. Department of Agricultiu*e. 

REFERENCES 

Alemdag, I. S. 1978. Evaluation of some competition indi- 
ces for the prediction of diameter increment in planted 
white spruce. Inf. Rep. FMR-X-108. Canadian Forest 
Service Forest Management Institute. 39 p. 

Arno, S. F. 1986. Whitebark pine cone crops — a diminish- 
ing source of wildlife food? Western Journal of Applied 
Forestry. 1(3): 92-97. 

Arno, S. F.; Weaver, T. 1990. Whitebark pine community 
types and their patterns on the landscape. In: Schmidt, 
Wyman C; McDonald, Kathy J., comps. Proceedings — 
symposium on whitebark pine ecosystems: ecology 
and management of a high-mountain resource; 1989 
October 29-31; Bozeman, MT. Gen. Tech. Rep. INT-270. 
Ogden, UT: U.S. Department of Agriculture, Forest Ser- 
vice, Intermountain Research Station: 97-105. 

Craighead, J. J.; Svunner, J. S.; Scaggs, G. B. 1982. A de- 
finitive system for analysis of grizzly bear habitat and 
other wilderness resources. Wildlands Inst. Monogr. 1. 
Missoula, MT: University of Montana. 279 p. 

Daniels, R. F. 1976. Simple competition indices and their 
correlation with annual loblolly pine tree growth. Forest 
Science. 22: 454-456. 

Daniels, R. F.; Burkhart, H. E.; Clason, T. R. 1986. A 
comparison of competition measures for predicting 
growth of loblolly pine trees. Canadian Journal of For- 
est Research. 16: 1230-1237. 



Eggers, D. E. 1986. Management of whitebark pine as po- 
tential grizzly bear habitat. In: Contreras, Glen P.; 
Evans, Keith E., comps. Proceedings — grizzly bear habi- 
tat symposiimi; 1985 April 30-May 2; Missoula, MT. 
Gen. Tech. Rep. INT-207. Ogden, UT: U.S. Department 
of Agriculture, Forest Service, Intermountain Research 
Station: 170-175. 

Eggers, D. E. 1990. Silvicultiiral management alterna- 
tives for whitebark pine. In: Schmidt, Wyman C; 
McDonald, Kathy J., comps. Proceedings — symposium 
on whitebark pine ecosystems: ecology and maneige- 
ment of a high-mountain resource; 1989 October 29-31; 
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Research Station: 324-328. 

Hamilton, G. J. 1969. The dependence of volume incre- 
ment of individual trees on dominance, crown dimen- 
sions, and competition. Forestry. 42: 133-144. 

Hansen-Bristow, K.; Montague, C; Schmid, G. 1990. 
Geology, geomorphology, and soils within whitebark 
pine ecosystems. In: Schmidt, W3anan C; McDonald, 
Kathy J., comps. Proceedings — symposium on 
whitebark pine ecosystems: ecology and management of 
a high-mountain resource; 1989 October 29-31; 
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Research Station: 62-71. 

Hegyi, F. 1974. A simulation model for managing jack- 
pine stands. In: Fries, J., ed. Growth models for tree 
and stand simulation. Res. Notes 30. Stockholm: Royal 
College of Forestry: 74-90. 

Kendall, K. C. 1983. Use of pine nuts by black and grizzly 
bears in the Yellowstone area. In: Bears — their biology 
and management; International Conference on Bear 
Research and Management. Madison, WI: International 
Bears Association. 5: 166-173. 

Kendall, K. C; Arno, S. F. 1990. Whitebark pine— an 
important but endangered wildlife resource. In: 
Schmidt, Wyman C; McDonald, Kathy J., comps. 
Proceedings — symposiima on whitebark pine ecosys- 
tems: ecology and management of a high-mountain 
resource; 1989 October 29-31; Bozeman, MT. Gen. Tech. 
Rep. INT-270. Ogden, UT: U.S. Department of Agricul- 
ture, Forest Service, Intermountain Research Station: 
264-273. 

Kipfer, T. R. 1992. Post-logging stand characteristics and 
crown development of whitebark pine (Pinus albicau- 
lis). Bozeman, MT: Montana State University, Depart- 
ment of Earth Science. 83 p. Thesis. 

Lorimer, C. G. 1983. Tests of age-independent competition 
indices for individual trees in natural hardwood stands. 
Forest Ecology and Management. 6: 343-360. 

Marston, R. A.; Anderson, J. E. 1991. Watersheds and 
vegetation of the Greater Yellowstone Ecosystem. Con- 
servation Biology. 5(3): 338-346. 

McCaughey, W. W.; Schmidt, W. C. 1990. Autecology of 
whitebark pine. In: Schmidt, Wyman C; McDonald, 
Kathy J., comps. Proceedings — symposivun on 
whitebark pine ecosystems: ecology and management 
of a high-mountain resource; 1989 October 29-31; 



134 



Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Research Station: 85-96. 

Powers, T. M. 1991. Ecosystem preservation and the 
economy in the Greater Yellowstone Area. Conservation 
Biology. 5(3): 395-404. 

Schmidt, W. C.; McCaughey, W. W. 1990. Whitebark 
pine — a subalpine species needing silvicultural atten- 
tion. In: Schmidt, Wyman C; McDonald, Kathy J., 
comps. Proceedings — symposium on whitebark pine eco- 
systems: ecology and management of a high-mountain 
resource; 1989 October 29-31; Bozeman, MT. Gen. Tech. 
Rep. INT-270. Ogden, UT: U.S. Department of Agricul- 
tiu-e, Forest Service, Intermountain Research Station: 
373-374. 

Spurr, S. H.; Barnes, B. V. 1980. Forest ecology. 3d ed. 
New York: John Wiley and Sons. 687 p. 

Tome, M.; Burkhart, H. E. 1989. Distance-dependent com- 
petition measures for predicting growth of individual 
trees. Forest Science. 35(3): 816-831. 

Wagner, R. G.; Radosevich, S. R. 1991a. Neighborhood 
predictors of interspecific competition in young 



Douglas-fir plantations. Canadian Journal of Forest Re- 
search. 21: 821-828. 
Wagner, R. G.; Radosevich, S. R. 1991b. Interspecific com- 
petition indices and other factors influencing the perfor- 
mance of Douglas-fir saplings in the Oregon Coast 
Range. Canadian Journal of Forest Research. 21: 829- 
835. 

Weaver, T.; Forcella, F.; Dale, D. 1986. Cone production 
in Pinus albicaulis forests. In: Shearer, R. C, comp. 
Proceedings — conifer tree seed in the inland mountain 
west symposium; 1985 August 5-6; Missoula, MT. Gen. 
Tech. Rep. INT-203. Ogden, UT: U.S. Department of 
Agriciilture, Forest Service, Intermotmtain Research 
Station: 68-76. 

Weaver, T.; Forcella, F.; Dale, D. 1990. Stand develop- 
ment in whitebark pine woodlands. In: Schmidt, 
Wyman C; McDonald, Kathy J., comps. Proceedings — 
symposimn on whitebark pine ecosystems: ecology 
and management of a high-mountain resource; 1989 
October 29-31; Bozeman, MT. Gen. Tech. Rep. INT-270. 
Ogden, UT: U.S. Department of Agricvilture, Forest Ser- 
vice, Intermountain Research Station: 151-155. 



135 



FIRE ECOLOGY OF WHITEBARK PINE 
FORESTS OF THE NORTHERN ROCKY 
MOUNTAINS, U.SA. 

Penelope Morgan 
Stephen C. Bunting 
Robert E. Keane 
Stephen F. Amo 



Abstract— Fires once occurred at intervals between 30 and 300 
years in whitebark pine (Pinus albicaulis) forests in the Northern 
Rocky Mountains, U.S.A., but since the early 1900's fewer fires 
have occurred, contributing to declining abundance of whitebark 
pine. In the absence of fire or other major disturbance, whitebark 
pine is replaced by other conifers on most of the upper subalpine 
landscape. Whitebark pines often survive low-intensity surface 
fires. Large stand-replacement fires also benefit this species by 
creating the open, burned sites where regeneration is most 
successful. 



In the Northern Rocky Mountains of western North 
America, whitebark pine (Pinus albicaulis) historically 
dominated many upper subalpine forests. These high- 
elevation forests usually have poorly developed, rocky 
soils and are often located within wilderness or roadless 
areas. As a consequence, whitebark pine is seldom har- 
vested for forest products, but it is important for scenic, 
watershed, and wildlife habitat values. 

Whitebark pine dominates middle- and late-successional 
stages. In the absence of major disturbance, however, 
whitebark pine is eventually replaced by the more shade- 
tolerant subalpine fir (Abies lasiocarpa) and Engelmann 
spruce (Picea engelmannii) in most of its range in the upper 
subalpine forest zone of the Northern Rocky Mountains 
(fig. 1). Whitebark pine is a common serai component 
of upper subalpine forests found on the Abies lasiocarpa/ 
Vaccinium scoparium, A. lasiocarpa/Luzula hitchcockii, 
and A. lasiocarpa/Arnica cordifolia habitat types (Pfister 
and others 1977; Steele and others 1981, 1983; Weaver 
and Dale 1974). These sites are cold, with July mean 
temperatures averaging 13 to 15 °C (Pfister and others 
1977). Although annual precipitation averages 610 to 
1,780 mm, summer drought is common (Arno and Hoff 
1989; Pfister and others 1977; Weaver and Dale 1974). 



Whitebark pine is also found in pure stands on relatively 
dry and severe, windswept sites near timberline where it 
is the climax tree species (Amo and HofF 1989). It is the 
sole climax tree species on Pinus albicaulis habitat types 
in Montana, central Idaho, and western Wyoming, and in 
southern Canada (Amo and Hoff 1989; Steele and others 
1981, 1983). Whitebark pine and subalpine fir are climax 
codominants on the Pinus albicaulis-Abies lasiocarpa habi- 
tat types where subalpine fir growth is stunted in the severe 
microclimate (Pfister and others 1977). Climax whitebark 
pine forests are usually open, with smcdl patches of trees 
of mixed ages interspersed with meadows (fig. 2). Average 
July mean temperatures are 10 to 12 °C with severe sum- 
mer droughts and frosts; the annual precipitation of 71 to 
153 cm falls mostly as snow (Arno and Hoff 1989; Pfister 
and others 1977). 

Fires are very important to regeneration and survival of 
whitebark pine on sites where it is serai. Whitebark pine 
often survives these low-intensity surface fires, which more 
easily kill associated conifers (fig. 3). Stand-replacing fires 
also benefit whitebark pine, for although all trees are usu- 
ally killed, whitebark pine regenerates on burned sites more 
successfully than many associated tree species (Tomback 




Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Penelope Morgan is Associate Professor, Department of Forest Re- 
sources, University of Idaho, Moscow, ID 83843; Stephen C. Bunting is 
Professor, Department of Range Resources, University of Idaho, Moscow, 
ID 83843; Robert E. Keane and Stephen F. Amo are Research Ecologist 
and Research Forester, Intermountain Fire Sciences Laboratory, Inter- 
mountain Research Station, Forest Service, U.S. Department of Agricul- 
ture, P.O. Box 8089, Missoula, MT 59801. 




Figure 1 — Subalpine fir and Engelmann spruce 
now dominate many stands and whole landscapes 
where whitebark pine was historically abundant. 
Whitebark pine is declining in abundance even 
where blister rust is uncommon, such as on this 
site east of Yellowstone National Park in Wyoming. 



136 



and others 1990) (fig. 4). Large stand-replacing fires are 
infi-equent, usually occurring only during windy conditions 
after prolonged drought. The fires that burned in and 
around Yellowstone National Park in 1988 were spectacu- 
lar examples of the large, high-intensity fires that periodi- 
cally burn within whitebark pine and adjacent forest types. 
These fires burned whitebark pine habitats in a patchy, 
stand-replacing manner. 

During the 10,000 years since development of forests af- 
ter the last glacial retreat, fires have had a major influence 
on the structure and composition of forests in the Northern 
Rocky Mountains (Arno 1980). Fire occurrence has been 
significantly altered by human activity. Native Americans 
used fire to manipulate vegetation, for hunting, for commu- 
nication, and for other purposes (Pyne 1982). More recently, 
humans have affected fire occiurence by purposefully or in- 
advertently igniting fires, through fire suppression, and by 
grazing domestic livestock, logging, or otherwise altering 
the fuels available to burn. Efi"orts to suppress fires have 
become increasingly effective since about 1935 (Arno 1980; 
Pyne 1982). 




Figure 3 — Whitebark pine trees often survive fires and individual trees may survive multiple 
fires. Information on fire frequency and effects is derived from dates of fire scars on tree 
sections such as this one from a 208-year-old whitebark pine tree in the Shoshone National 
Forest in northwestern Wyoming. First scarred by fires when only 12.2 cm in diameter at 
the base, it was again scarred 60 years later. It was cut in 1 988, 1 1 0 years later. 




Figure 2 — Whitebark pine often is the only tree 
that can grow on harsh sites. Here, in the Challis 
National Forest in central Idaho, whitebark pine is 
the climax tree species. 



137 




Figure 4 — Young whitebark pine trees regener- 
ated successfully following a stand-replacement 
fire in northwestern Wyoming. All trees were killed 
by the fire, which occurred 55 years before this 
photograph was taken in 1 988. 



WHITEBARK PINE DECLINE 

Whitebark pine has decHned in abundance in major por- 
tions of its range (Arno 1986). Once important on 10 to 15 
percent of the forested landscape in the Northern Rocky 
Mountains (Arno and Hoff 1989), whitebark pine mortality 
rates averaged 42 percent over the last 20 years in western 
Montana (Keane and Arno 1993). Arno and others (1993) 
found that for a 200-ha study area in Montana, the per- 
centage of stands with at least 20 percent basal area of 
mature whitebark pine declined from 37 percent in 1900 
to 20 percent in 1991; 14 percent of the area was dominated 
by whitebark pine in 1900; none was so dominated in 1991. 

Whitebark pine decline is most pronounced on the more 
productive sites where subalpine fir and Engelmann spruce 
are highly competitive (Arno 1986; Ciesla and Fumiss 1986; 
Keane and Arno 1993; Kendall and Arno 1990). Declining 
whitebark pine threatens wildlife habitat because the seeds 
of whitebark pine are a valuable food for many birds and 
small mammals, including the endangered grizzly bear 
(Kendall and Arno 1990). 

Recent decline in whitebark pine abundance is linked 
to less frequent fires (Keane and Morgan, these proceed- 
ings; Keane and others 1990). Fires in whitebark pine for- 
ests occurred at mean intervals of 30 to 300 years based on 
fire history information derived from fire scars and stand 
ages (table 1). Fewer fires in the last 50 to 100 years (Arno 
and Hoff 1989; Morgan and Bunting 1990) have resulted 
in extensive changes in the composition of forests in the 
high-elevation landscapes of the Northern Rocky Mountains, 
Composition of subalpine forests has shifted dramatically 
toward dominance by subalpine fir and Engelmann spruce 
(Keane and others 1993). The decline of whitebark pine 
has been further exacerbated by the introduced blister rust 
{Cronartium ribicola) and the native mountain pine beetle 
(Dendroctonus ponderosae), both of which kill whitebark 
pine but not subalpine fir or Engelmann spruce. 



STAND-REPLACING FIRES 

In conditions of extreme drought lasting more than 
2 years, fires ignited by lightning and fanned by high winds 
can rapidly spread and kill trees in large patches. These 
fires usually bum in other forest types as well, converting 
large segments of the landscape to early successional plant 
communities. Fires that spread through forests at lower 
elevations historically burned into the adjacent whitebark 
pine forests (Arno and Hoff 1989). In whitebark pine for- 
ests, stand-replacing fires typically spread on the ground 
(Lasko 1990). Fires may kill trees by scorching foliage or 
by heating the bole or roots to lethal temperatures. Some- 
times, crown fires occiu- that bum through the tree crowns, 
killing all trees in their paths (Lasko 1990). 

Stand-replacing fires provide important opportunities 
for whitebark pine to regenerate. Many competing tree 
species rely on the wind to disseminate seed. Whitebark 
pine has a distinct advantage in regenerating following 
extensive disturbances (Tomback and others 1990). The 
Clark's nutcracker {Nucifraga columbiana) conmionly trans- 
ports seeds several kilometers (Hutchins and Lanner 1982). 
These birds prefer open, burned areas for caching seeds 
(Tomback and others 1990). Thus, although large fires are 
infrequent, they are ecologically important in maintaining 
extensive whitebark pine forests on the landscape. 

LOW-INTENSITY SURFACE FIRES 

Low-intensity surface fires also influence the relative 
abundance of whitebark pine on the landscape. Such 
fires are more frequent and smaller in extent than stand- 
replacing fires. Low-intensity fires generally kill young 
whitebark pine and both large and small subalpine fir. 
Such fires can result in open, parklike stands of nearly 
pure whitebark pine (Arno 1986). Some fires probably 
burned as low-intensity surface fires but later became 
stand-replacing fires when burning conditions were more 
severe. That whitebark pine trees often survive surface 
fires is evidenced by the many living trees that have scars 
from one or more fires (fig. 1). 



Table 1— Fire frequency from whitebark pine forests expressed 
as the mean and range (in parentheses) of the years 
between fires 



Fire frequency 


Geographical area and reference 


144 (55 to 304) 


Bob Marshall Wilderness Complex, north- 




western fy/lontana (Keane and others 1 993) 


80 (50 to 300) 


Bitterroot Mountains, Montana (Arno 1980) 


30 to 41 (4 to 78) 


1 00- to 300-ha stands where subalpine fir 




is climax, Montana (Arno 1986) 


29 (13 to 46) 


10 stands within 100 ha, northwestern 




Wyoming (Morgan and Bunting 1990) 


300 


Lodgepole pine forests adjacent to but 




at lower elevations than whitebark pine. 




Yellowstone National Park (Romme 1 982) 



138 



These low-intensity fires are more common on relatively 
dry sites, occurring only where stand structures, fuel ac- 
cumulation, and microclimatic conditions are conducive. 
Thus, such fires result in many small burned patches, in- 
creasing landscape heterogeneity. 

Where fires are more fi'equent, they are more likely to be 
of low intensity. Morgan and Bimting (1990) documented 
very fi'equent low-intensity fires on a relatively dry site 
supporting serai whitebark pine in open, parkHke stands 
within a 100-ha area in northwestern Wyoming. There the 
mean interval between fires was 33 years prior to 1867. 
Fires were much more common prior to 1850 than they 
have been since then (Morgan and Bunting 1990). 

FIRE REGIMES 

Through time, most whitebark pine forests experience a 
mixture of stand-replacement and low-intensity fires. The 
frequency of these types of fires within a given landscape 
will vary with landscape complexity and heterogeneity. 
Stand-replacing fires are more common diiring regional 
droughts, and often burn large patches regardless of fuel 
loading or stand condition. Fire behavior and effects are 
also influenced by the stand structure and fuel accumula- 
tion, which are in part determined by the time since last 
bum. 

Where whitebark pine is cHmax, fires are infrequent and 
generally of low intensity. In whitebark pine krummholz 
and ribbon forests, fires are infrequent and of variable in- 
tensity. When fires do occ\ir, many trees die and regenera- 
tion is very slow. Keane and others (1990) predict where 
blister rust infection rates are high, climax whitebark pine 
forests will convert to herbaceous or shrub commtmities 
following fire. 

Stand-replacing fires are more common where whitebark 
pine is a serai dominant. Stand-replacing fires become in- 
creasingly likely with advancing succession (Fischer and 
Clayton 1983; Morgan and Bimting 1990). 

SUCCESSION FOLLOWING FIRE 

Fire is a key process affecting serai whitebark pine forest 
structure sind composition. Successional patterns on sites 
where whitebark pine is serai are predictable (fig. 5), but 
they are not closely tied to stand age or time since last dis- 
tiu-bance (Mattson and Reinhart 1990). The stand struc- 
ture and the microsites created vary from fire to fire. Con- 
ditions for successful regeneration of tree seedlings are 
sporadic, depending on favorable climatic and site condi- 
tions. Althoiigh both subalpine fir and whitebark pine 
may establish soon after a fire, it may take a half-century 
or longer for a forest to develop. 

Whitebark pine is one of the first tree species to become 
established in abimdance following stand-replacing fires 
(Weaver and Dale 1974). As a consequence, it often domi- 
nates initially, often for up to 225 years or more (Loope 
and Gruell 1973; Morgan and Bimting 1990). Early serai 
stands are dominated by whitebark pine seedlings and sap- 
lings growing along with a dense herbaceous and shrub 
understory. Subalpine fir seedlings are often present, 
especially close to parent trees that survived the fire, but 
they grow more slowly than whitebark pine trees (Arno 



Late Serai 
(ABLA, PEN, 
PIAL, PICO) 




Shrub/herb 




Late Mid-Seral 
(ABLA, PIAL, 
PICO, PEN) 




Early Sera! 
(PIAL, PICO 
ABLA, PEN) 



Mid-Serai 
(PIAL, PICO, 
ABLA, PEN) 



^^^^^^ 



Stand-replacing fire 
Return to shrub/herb stage 



Low-intensity fire 
Many PIAL survive 



Figure 5 — Generalized forest succession following 
fires on sites where whitebark pine (PIAL) is re- 
placed by subalpine fir (ABLA) and Engelmann 
spruce (PEN) with advancing succession. Lodge- 
pole pine (PICO) is a common associate. Low- 
intensity surface fires may occur at any stage 
but are most likely in mid-seral stands. Stand- 
replacing fires are the norm in late-seral stands. 
Tree species listed in order of abundance. 
Adapted from Fischer and Clayton (1983). 



and Hoff 1989). Whitebark pine seedlings are more abun- 
dant than subalpine fir seedlings in large burned areas 
because whitebark pine seedlings are dispersed farther by 
the Clark's nutcracker. Subalpine fir seeds are dispersed 
by the wind. Many standing snags and fallen logs are 
present. Most whitebark pine trees do not produce large 
numbers of seeds imtil at least age 70; most do not produce 
any cones xmtil age 50 (Morgan and Bvmting 1992). With 
time, whitebark pine is gradually replaced by subalpine fir 
and Engelmann spruce. In late-seral stands, both the tree 
canopy and the imderstory are dominated by many subal- 
pine fir trees. 

MANAGEMENT IMPLICATIONS 

Continued decline in whitebark pine abundance threat- 
ens to dramatically reduce the availability of seeds for the 
many animals that rely on them as a food source (Arno 
1986). 

Fire exclusion greatly reduces opportimities for regen- 
eration of whitebark pine. Cone production is higher in 
the stands where whitebark pine is healthy and dominant 
(Morgan and Bunting 1992). With blister rust reducing 
cone production and killing parent trees (Arno and Hoff 
1989), the seed available for tree regeneration is rapidly 



139 



declining (Amo 1986; Keane and Amo 1993). As cone pro- 
duction declines, animals eat more of the seeds, leaving 
fewer to regenerate. Whitebark pine cone production de- 
clines with advancing succession (Morgan and Bunting 
1992) and as infection by blister rust increases (Arno and 
Hoff 1989). Given the rapidly declining abundance of white- 
bark pine in some regions, we must act quickly to create op- 
portunities for individuals that are resistant to blister rust 
to regenerate. Although whitebark pine is very sensitive 
to blister rust, some individual trees are genetically resis- 
tant to infection (Hoff and others 1990). If we do not create 
opportunities for those trees to regenerate before their cone 
production declines or they die in advancing succession, 
opportunities for enhancing natural mechanisms of white- 
bark pine recovery will be lost. 

Three options are available for improving the health and 
productivity of whitebark pine stands that are now domi- 
nated by subalpine fir and spruce. One option is creating 
forest openings through timber harvest, girdling, or other- 
wise killing trees mechanically. Cutting trees to create 
openings to encourage regeneration of whitebark pine is 
possible but may not be economically feasible. Although 
there are often small whitebark pine trees growing in the 
imderstory of mixed conifer stands, they are often very old 
and of poor vigor and are therefore unlikely to respond 
when larger trees are removed. 

Another option is the liberal use of prescribed fire to cre- 
ate openings for regeneration and to favor whitebark pine 
in stands now codominated by subalpine fir. Managers can 
purposefully ignite fires or allow lightning fires to burn 
under carefully prescribed conditions of weather, fuel, and 
location. Historically, fire was the primary natural distur- 
bance, and it may be the most practical tool for managing 
whitebark pine considering economics, policy, and topo- 
graphic limitations for using timber harvest. 

A third option, to use a combination of techniques, may 
be most successful. Arno and Keane are involved in an ef- 
fort to test alternative techniques for perpetuating white- 
bark pine in a blister rust-infected area of the Bitterroot 
National Forest in Montana. 

ACKNOWLEDGMENTS 

Research funding was provided through the USDA Forest 
Service, Intermountain Research Station. This is Contribu- 
tion No. 693 of the Idaho Forest, Wildlife and Range Experi- 
ment Station, University of Idaho, Moscow, ID 83843. 
We thank Bill Fischer and Elizabeth Reinhardt for their 
comments. 

REFERENCES 

Arno, S. 1980. Forest fire history in the Northern Rockies. 
Journal of Forestry. 78(8): 460-465. 

Arno, S. 1986. Whitebark pine cone crops — a diminishing 
source of wildlife food? Western Journal of Applied For- 
estry. 1(3): 92-94. 

Arno, S. F.; Hoff, R. J. 1989. Silvics of whitebark pine 
{Pinus albicaulis). Gen. Tech. Rep. INT-253. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Research Station, lip. 



Arno, S. F.; Reinhardt, E. D.; Scott, J. H. 1993. Forest 
structure and landscape patterns in the subalpine lodge- 
pole pine t)^es: a procedure for quantifying past and 
present conditions. Gen. Tech. Rep. INT-294. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, 
Intermountain Research Station. 17 p. 

Ciesla, W. M.; Furniss, M. M. 1986. Idaho's haimted forests. 
American Forests. 81(8): 23-25. 

Fischer, W. C; Clayton, B. D. 1983. Fire ecology of Montana 
forest habitat types east of the Continental Divide. Gen. 
Tech. Rep. INT-141. Ogden, UT: U.S. Department of 
Agriculture, Forest Service, Intermountain Forest and 
Range Experiment Station. 83 p. 

Hoff, R. J.; Bingham, R. T.; McDonald, G. I. 1980. Relative 
blister rust resistance of white pines. European Journal 
of Forest Pathology. 10: 307-316. 

Hutchins, H. E.; Lanner, R. M. 1982. The central role of 
Clark's nutcracker in the dispersal and establishment 
of whitebark pine. Oecologia. 55: 192-201. 

Keane, R. E.; Arno, S. F. 1993. Rapid decline of whitebark 
pine in western Montana: evidence from 10-year meas- 
urements. Western Journal of Applied Forestry. 8(2): 
44-47. 

Keane, R. E.; Arno, S. F.; Brown, J. K; Tomback, D. F. 
1990. Modelling stand dynamics in whitebark pine 
forests. Ecological Modelling. 51: 73-95. 

Keane, R. E.; Morgan, P.; Menakis, J. 1993. Landscape 
assessment of the decline of Pinus albicaulis in the 
Bob Marshall Wilderness Complex, Montana, USA. 
Missoula, MT: U.S. Department of Agriculture, Forest 
Service, Intermountain Research Station, Intermountain 
Fire Sciences Laboratory. Review draft. 

Kendall, K C; Arno, S. F. 1990. Whitebark pine — an impor- 
tant but endangered wildlife resource. In: Schmidt, W. C; 
McDonald, K. J., comps. Proceedings — sjmaposium on 
whitebark pine ecosystems: ecology and management of 
a high-mountain resource; 1989 March 29-31; Bozeman, 
MT. Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Depart- 
ment of Agriculture, Forest Service, Intermountain Re- 
search Station: 264-273. 

Lasko, Richard J. 1990. Fire behavior characteristics and 
management implications in whitebark pine ecosystems. 
In: Schmidt, W. C.; McDonald, K. J., comps. Proceedings — 
sjmiposium on whitebark pine ecosystems: ecology and 
management of a high-mountain resource; 1989 March 
29-31; Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, 
UT: U.S. Department of Agriculture, Forest Service, In- 
termountain Research Station: 319-323. 

Loope, L. L.; Gruell, G. E. 1973. The ecological role of fire 
in the Jackson Hole area, northwestern Wyoming. Qua- 
ternary Research. 3: 425-443. 

Mattson, D. J.; Reinhart, D. P. 1990. Whitebark pine on the 
Mount Washburn Massif, Yellowstone National Park. In: 
Schmidt, W. C; McDonald, K. J., comps. Proceedings- 
symposium on whitebark pine ecosystems: ecology and 
management of a high-mountain resource; 1989 March 
29-31; Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, 
UT: U.S. Department of Agriculture, Forest Service, In- 
termountain Research Station: 106-117. 

Morgan, P.; Bunting, S. C. 1990. Fire effects in whitebark 
pine forests. In: Schmidt, W. C; McDonald, K. J., comps. 



140 



Proceedings — symposium on whitebark pine ecosystems: 
ecology and management of a high-momitain resource; 
1989 March 29-31; Bozeman, MT. Gen. Tech. Rep. 
INT-270. Ogden, UT: U.S. Department of Agriculture, 
Forest Service, Intermoimtain Research Station: 166-170. 

Morgan, P.; Bunting, S. C. 1992. Cone crops in whitebark 
pine (Pinus albicaulis) forests in the Greater Yellowstone 
Area. Moscow, ID: University of Idaho, College of For- 
estry, Wildlife and Range Sciences. 21 p. Review draft. 

Pfister, R. D.; Kovalchik, B. L.; Arno, S. F.; Presby, R. C. 
1977. Forest habitat types of Montana. Gen. Tech. Rep. 
INT-34. Ogden, UT: U.S. Department of Agriculture, 
Forest Service, Intermountain Forest and Range Experi- 
ment Station. 174 p. 

Pyne, S. J. 1982. Fire in America. Princeton, NJ: Princeton 
University Press. 654 p. 

Romme, W. H. 1982. Fire and landscape diversity in subal- 
pine forests of Yellowstone National Park. Ecological 
Monographs. 52: 199-221. 

Steele, R.; Pfister, R. D.; Ryker, R. A.; Kittams, J. A. 1981. 
Forest habitat types of central Idaho. Gen. Tech. Rep. 



INT-114. Ogden, UT: U.S. Department of Agriculture, 
Forest Service, Intermountain Forest and Range Ex- 
periment Station. 138 p. 

Steele, R.; Cooper, S. V.; Ondov, D. M.; Roberts, D.; 

Pfister, R. D. 1983. Forest habitat types of eastern Idaho- 
western Wyoming. Gen. Tech. Rep. INT- 144. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Forest and Range Experiment Station. 122 p. 

Tomback, D. F.; Hoffman, L. A.; Sund, S. K. 1990. Coevo- 
lution of whitebark pine and nutcrackers: implications 
for forest regeneration. In: Schmidt, W. C; McDonald, 
K. J., comps. Proceedings — symposium on whitebark 
pine ecosystems: ecology and management of a high- 
mountain resource; 1989 March 29-31; Bozeman, MT. 
Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Department 
of Agriculture, Forest Service, Intermountain Research 
Station: 118-130. 

Weaver, T.; Dale, D. 1974. Pinus albicaulis in central 
Montana: environment, vegetation and production. 
American Midland NaturaHst. 92: 222-230. 



141 



VEGETATION DISTRIBUTION AND 
PRODUCTION IN ROCKY MOUNTAIN 
CLIMATES— WITH EMPHASIS ON 
WHITEBARK PINE 

Tad Weaver 



Abstract — The distribution and production of vegetation on the 
altitudinal gradient (grassland-forest-alpine) was plotted against 
climatic parameters to evaluate hypothetical controlling factors. 
(1) Whitebark pine (Pinus albicaulis) is likely excluded from 
higher zones by a cool growing season or wind-induced drought. 
It is probably not excluded by low temperatures occurring during 
its hardening, hard, or dehardening seasons. (2) While the lower 
physiological limit of whitebark pine is probably set by drought, 
its lower realized limit is directly set by subalpine fir (Abies lasio- 
carpa) and lodgepole pine (Pinus contorta) competitors and indi- 
rectly set by factors that control their distribution. (3) The upper 
limits for most other dominant species are probably set by grow- 
ing season temperature. The lower Umits are likely set by com- 
petition down to the cedar-hemlock (Thuja plicata I Tsuga hetero- 
phylld) zone and by drought in drier areas. (4) Production is 
strongly correlated (r^ = 0.86) with growing season length (soil 
thawed season minus dry soil days). Multiplying season length 
by average temperature did not improve the growing season pre- 
dictor, perhaps because vegetation at each altitude is especially 
adapted to temperatures in its zone. 



Vegetation composition and structure vary along gradi- 
ents of temperature and precipitation v^^hether the condi- 
tion changes with altitude (Daubenmire 1956; Gams 1931; 
Weaver 1980) or geography (Holdridge 1967; Walter 1973; 
Whittaker 1975). Graphical devices based on simple pa- 
rameters such as average annual temperature and total 
annual precipitation (Holdridge 1967; Whittaker 1975) 
are workable predictors of the vegetation growing in par- 
ticular climates. It seems obvious, however, that these 
devices succeed, not because the climatic parameters used 
are causal, but because the parameters are correlated 
with causal climatic factors. 

Production of vegetation is largely determined by cli- 
mate and is therefore expected to be highest where tem- 
perature, moisture, and nutrients are simidtaneously 
favorable. In Rocky Mountain vegetation one therefore 
expects production to be highest in low-altitude (warm) 
moist forests (Thuja-Tsuga) and to decline both with in- 
creasing altitude (because of cooling and reduced nutri- 
ent availability) (Weaver 1979) and with decreasing alti- 
tude (because of decreasing water availability). While it is 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Tad Weaver is Plant Ecologist, Biology Department, Montana State 
University, Bozeman, MT 59717. 



rarely done, vegetation production should also be success- 
fully plotted in a Hutchinsonian (1958) hyperspace with 
axes that are either physiologically meaningful or surro- 
gates well correlated with physiologically meaningful axes. 

The object of this paper is to correlate vegetation 
performance — survival and production — ^with physiologi- 
cally meaningftd aspects of temperature and moisture. 
Study of the distribution of species on these prestunptively 
causal axes will eliminate some hypotheses of cause and 
sharpen others for experimental tests. While whitebark 
pine (Pinus albicaulis) is the primary subject of this paper, 
the approach could be applied to other species, as well. 

METHODS 

Impacts of climatic factors on vegetation distribution 
were evaluated by comparing factor levels among seg- 
ments on a vegetational gradient. In essence, the method 
allows one to deduce (1) that a presmnptive factor that 
does not vary between vegetation t3T)es is not controlling 
the vegetation changes observed, and (2) that a presimip- 
tive factor that does change between vegetation types de- 
serves further discussion as directly controlhng, indirectly 
controlling, or a noncausal correlate. The method is illus- 
trated here with discussion of whitebark pine, but it could 
be applied to many other species as well. 

The Data 

Environmental zones were identified by climax vegeta- 
tion occupying them (Daubenmire 1943; Daubenmire and 
Daubenmire 1968; Huschle and Hironaka 1980). They in- 
cluded, from low to high altitude, desert shrub, dry grass- 
land, warm forest, cool forest, and alpine ecosystems. 
Specific zones are listed with their Kuchler (1964) type 
numbers and representative stations in table 1. 

Climates of the environmental zones were characterized 
by using climatic statistics from approximately five sta- 
tions in each environmental-vegetation zone. To maintain 
the integrity of the data, raw data are plotted wherever 
possible. When necessary, medians were used to repre- 
sent "typical" sites; use of medians deemphasizes sites 
vrith conditions intermediate between modal conditions 
of adjacent types. The stations chosen include most avail- 
able stations and may thus be considered as a "complete 
sample." The sample may not be entirely random because 
weather stations must be accessible; in higher types, for 
example, the sites may be relatively low. 



142 



Table 1 — Weather stations representative of major Northern Rocky Mountain environmental zones. ^ Zones are listed from high to low altitude 



Zone Station locations 



Alpine^ 

Whitebark pine^ 
Subalpine fir 
Douglas-fir 
Cedar-hemlock 
Ponderosa pine 

Idaho fescue 
Bluebunch 
wheatgrass 
Grama grass 
Atriplex spp. 



White Mountain, CA; Niwot Ridge, CO 

Old Glory, BC; Kings Hill, MT; Ellery Lake, CA; Crater Lake, OR 

High: Summit, MT; West Yellowstone, MT; Lake Yellowstone, MT; Low: Burke, ID; Seeley Lake, MT; Hungry Horse, MT 
East: Hebgen Lake, MT; Lakeview, MT; Lamar, WY; Palisades Dam, ID; Dixie, ID; West: Libby, MT; Lincoln, MT 
West Glacier, MT; Sandpoint, ID; Pierce, ID; Avery, ID; Priest, ID 

West: Garden Valley, ID; Poison, MT; Potlatch, ID; Kooskia, ID; East: Melstone, MT; Busby, MT; Colstrip, MT; Lame 
Deer, MT; Roundup, MT 

Gallatin Gateway, MT; Virginia City, MT; Mystic Lake, MT; Wisdom, MT; Bozeman MSU, MT; White Sulphur Springs, MT 

Browning, MT; Belgrade, MT; Kalispell, MT; Ennis, MT; Dillon, MT; Augusta, MT; Crow Agency, MT; Billings, MT 
Ekalaka, MT; Jordan, MT; Malta, MT; Rock Springs, MT; Rapelje, MT; Chester, MT; Great Falls, MT; Fairfield, MT 
Lovell, WY; Worland, WY; Basin, WY; Powell, WY; Deaver, WY 



'The vegetation types, with their Kuchler (1964) type numbers (KTXX), are: Deschampsia caespitosa-Carex (KT52), Pinus albicaulis (KT15), Abies lasiocarpa 
(KT1 5), Pseudotsuga menziesii (KT1 2), Thuja plicataJTsuga heterophylla (KT2), Pinus ponderosa (KT1 1,16), Festuca idatioensis (KT63), Agropyron spicatum 
(KT63), Bouteloua gracilis (KT64), Atripiex (KT40). Abies types from higher and lower altitudes, Pseudotsuga types from east and west of the mountains, and pon- 
derosa pine types from east and west of the Rockies are separately presented to emphasize possible differences. 

n"he White Mountain alpine site (1 0,1 50 ft) is unusually dry. The Ellery Lake whitebark pine site has only precipitation data; temperature data were therefore 
taken from a similar altitude at a nearby White Mountain site (9,645 ft). 



Killing Temperatures 

The hypothesis that extreme temperatiires might ex- 
clude a plant (for example, whitebark pine) from an adja- 
cent zone was tested by determining whether the condi- 
tion is more extreme in the iinoccupied than in the 
occupied zone; if it is not, the hypothesized factor is con- 
sidered unlikely to be lethal. This test was used on tem- 
peratures of early fall frosts, midwinter extreme lows, 
midwinter average lows, late fall frosts, midsummer aver- 
age highs, and midsummer extreme highs. Midwinter 
lows £ind midsvmimer highs were tested because these 
temperatures might exceed the potential tolerances of a 
plant; fall and spring frosts were considered because they 
might kill partially hardened plants. Fall frost tempera- 
tiu-es were measured as the absolute low in the month in 
which average temperature first fell below 0 °C; where no 
month had an average temperatvire of 0 °C, the values 
were interpolated to 0 °C from adjacent months. Spring 
frost temperatiu-es were estimated similarly for the 
month in which average temperatures first rose above 
0 °C. I argue later (see growing season) that average tem- 
peratures below 0 °C restrict water and nutrient uptake 
and therefore open and close the growing season; the use 
of 0 °C is in contrast to the 5 °C used by many phenolo- 
gists (for example, Chang 1968). "Absolute" temperatures 
were the lowest (or highest) seen in the period of record — 
10 years or more. 

Starvation Temperatiu*es 

The plant might be excluded from a zone free from kill- 
ing events if (temperatiire) conditions failed to support 
net photosynthesis on an annual basis. Two indices of 
growing season temperature were used to test the possi- 
bility of such starvation. First, temperatures were aver- 
aged across all growing season (defined later) months. 
Despite its common use in predicting growth (Chang 1968; 
Larcher 1975), this index is expected to underestimate the 



benefits of warmer temperatvires because chemical reaction 
rates increase exponentially with increasing temperature. 
An alternate index of temperature was therefore tested. 
In the second index temperatures were replaced with 
growth support units (gsu) and these were averaged 
across the season. On the assimiptions that Q^^ = 2, that 
very slow growth begins at 1 °C, and that native plants 
tolerate normal high temperatures of their zones, the 
growth support unit curve was constructed by interpolat- 
ing between 0 growth rate imits (gsu) at 0 °C, 1 gsu at 
1 °C, 2 gsu at 11 °C, 4 gsu at 21 °C, 8 gsu at 31 °C, and 
16 gsu at 41 °C. That is, we expect no growth at 0 °C or 
below, 1 vmit at 1 °C, 2 imits at 11 °C, 4 units at 21 °C, etc. 

Killing Drought 

Evaporation equals roughly 2 mm/°C x month 
(Daubenmire 1956; Nielsen 1986; Stephenson 1990; 
Thornthwaite 1948; Walter 1973). As a result, one might 
index drought duration by counting months when average 
precipitation (mm/2) is less than average temperature 
(C) or drought intensity by summing, across growing sea- 
son months, precipitation deficits below the calculated 
balance. 

A covmt of months registering drought on the Walter in- 
dex may overestimate drought duration where the deficit 
is small because drought may open late or close early in 
the month. To minimize this effect I have normalized 
drought duration from a scatter diagram of duration 
against deficit: 0-5 mm = 0 months, 6-15 mm = 1 month, 
16-35 mm = 2 months, 36-50 mm = 3 months, and 
51-85 mm = 4 months. 

I have used the simi of warm-season (average tempera- 
ture above 0 °C) water deficits as an index of drought 
magnitude. I do so with the recognition that, due to ex- 
clusion of important factors (for example, wind, Penman 
1949, and exponential temperature effects, Stephenson 
1990), this index usually underestimates water deficit. 



143 



Water Starvation 



RESULTS AND DISCUSSION 



"Water starvation" should occur where plants are not 
desiccated, but where water supplies do not support net 
photosynthesis on an annual basis. For example, starva- 
tion would occur if water were continually available, but 
evapotranspiration equaled or exceeded uptake rates. Such 
starvation would select for organisms with lower ET rates: 
Mesophytic leaves would be replaced by xeroph)i;ic leaves 
and leaf exposiu-e might be reduced by reduction of stature 
even to the point where plants existed only at the groimd 
surface or under transparent rock. 

Growing Season 

For sensitive introduced plants, frost-free season is of- 
ten recorded as an indicator of season length (Burke and 
others 1976; USGS 1971). 

Since cool region plants normally tolerate growing sea- 
son frosts, I argue for a different index. Growth requires 
both leaf activity (photos3nithesis) and root activity 
(water and nutrient absorbtion). For plants with frost- 
insensitive leaves (Burke and others 1976), midday tem- 
peratures should be high enough for photosynthesis long 
before soils are thawed. Root activity extends (maxi- 
mally) from spring soil thaw to the fall refreeze, so this 
period should be a better index of growing season length. 
Thus I open and close the growing season with 0 °C aver- 
age monthly air temperatures. So represented, initial ac- 
tivity in grasslands begins at spring soil temperatures 
(0-25 cm) near 0 °C and activity ceases in the fall with soil 
temperatures as high as 5 °C (Weaver, in preparation). 
In, or adjacent to, the subalpine fir {Abies lasiocarpa) zone 
the growing season may exceed this period because deep 
snow cover may prevent the freezing of soil water. 

While the temperature-bounded season probably ap- 
plies at high-altitude sites, at low-altitude sites soil dry- 
ing may close the growing season before frost does. Thus, 
one may better index growing season as warm season 
months minus any drought months included in the warm 
season. While the Northern Rocky Mountain drought sea- 
son comes at summer's end (Weaver 1980), this index will 
also apply to regions where the warm season opens with 
dry months. It is conceivable that other periodic factors, 
for example nutrients or pest attack, might affect growing 
season length. 

Production 

Production estimates for each series were drawn from 
the literature (table 2). These are correlated with climatic 
conditions that might control them: warm season length, 
warm moist season length, and the product of warm soil 
season length and temperature. For the latter, two tem- 
perature expressions were tested: average temperature 
minus 5 °C and Q (the Q^^-based average). The third 
(product) indices were expected to be best because they 
assume that plants grow when soils are unfrozen-moist 
and that their growth rate is proportional to temperature 
(over 5 °C for the first linear index and over 1 °C for the 
second exponential index). 



In the Northern Rocky Motmtains vegetation changes 
from grass-shrub (Atriplex, Bouteloua, Agropyron, and 
Festuca) to conifer {Pinus ponderosa, Thuja-Tsuga, 
Pseudotsuga, and Abies) to alpine graminoid with increas- 
ing altitude. Subzones listed parenthetically are described 
by Daubenmire (1943) and mapped by Kuchler (1964). 
This paper compsu-es weather data gathered in these zones 
to test h3TJothesized controls of distribution and production. 
The method is illustrated with whitebark pine. 

Upward Limits 

I speculate that the upward limits of whitebark pine 
might be determined by lethal factors (winter low tem- 
peratures, spring-fall fi*osts, or desiccation) or production 
deficits (starvation) due to inadequate water or heat 
units. These h3T)otheses are considered in the following 
paragraphs. 

If winter lows are lower in the alpine than in the pine 
zone below, winter cold might be the excluding factor, oth- 
erwise not. Contrary to my expectation, neither absolute 
lows nor average January minima are lower in the alpine 
than in the pine (or most other) vegetation zone(s) (fig. 1). 
The failure of low temperatures to develop at higher alti- 
tudes may be due to the high density of cold air: Cold air 
entering from the north stays low like mercury poured \m- 
der water and air cooled by exposure to cold alpine ground 
runs off (Geiger 1965). Since extreme temperatures are 
not lower at high than low altitude, I deduce that neither 
whitebark pine, nor most other native dominants, are ex- 
cluded from high sites by extreme low temperatures of win- 
ter. An exception to this generalization seems to appear in 
high- or low-altitude fi-ost pockets, where accumulating 
cold air may kill trees in winter or spring (Weaver 1990). 

For natives of an area, I suggest that frost damage is 
more likely when plants are partially hardened (fall) or 
partially dehardened (spring) than at midwinter. I expect 
(1) winter to be delimited approximately by the fall month 
in which surface soils fall below 0 °C and the spring 
month in which they rise above 0 °C, (2) winter to be a 
season of low root activity, low water, and low nutrient 
uptake, and (3) 0 °C soil temperatures to be approxi- 
mately coincident with 0 °C air temperatures (Weaver, in 
preparation). I therefore (1) compare, across vegetation 
types, observed temperature lows in these spring-fall sea- 
sons of incomplete frost hardness (fig. 2), (2) discover no 
difference, and (3) conclude that frosts probably do not 
partition vegetation native to the region. The preceding 
discussion depends on the assumption that plants harden 
and deharden in phase with air/soil temperatures. This 
argument would be fallacious if day length were the pri- 
mary controller of the hardening/dehardening process. 
However, since both temperature and correlated day 
length triggers are important (Salisbury and Ross 1992), 
I expect frosts to exclude plants from environmentally dis- 
tant, but not environmentally adjacent, vegetation types. 

While frosts and winter freezes probably do not exclude 
whitebark pine from the alpine (or other plants from the 
vegetation zone immediately above), a lack of heat imits 



144 



Table 2 — Productivity in grams per square meter per year and aboveground standing crop (Abvegr std crp) of major Rocky Mountain vegetation 
types 



Vegetation type 


Productivity^ 


AKuonr QtH f^m 






g/m^/yr 


t/ha (age.yr) 




Alpine 


135 


1.35 


Thilenlus and others 1974 




100-200 


1.00-2.00 


Scott and Billings 1 964 


Pinus albicaulis 


'60 


^140 ^300-500^ 


VVCdVd Cll IVJ L^ClIC' 1 / H 




'25-75 (53) 




PfiQtpr anH nthorQ 1Q77 

r IIOLC7I al IVJ \JlllC7IO 1*7/ 1 




200-700 


350 (300-500) 


Forcella and Weaver 1 977 


Abies lasiocarpa 


1 00-200 


160 f300-700) 


Aoipt and other*; 1QfiQ 




860 


357 M06 vr) 


Whitt^^kpr ^r\c\ Niprinn 1 07*^ 

V V 1 II llCir\C7l ai \\a I NIOI ll im I ^ r w 




'95-180 (137) 




Pfister and others 1 977 






100-250 (Old) 


Weaver and Forcella 1977 






150-250 


Landis and Mogren 1975 


Ps6udotsuga menziesii 


1,550 


438 (252) 


Whittakpr and Niprinn 1Q7^ 




'30-170 (150) 




Pfi^tpr and nthpr^ 1Q77 






100-350 (350+) 


Weaver and Forcella 1 977 


Thuja plicata- 








Tsuaa hsteroohvlla 


870 




Hanlfiv 1 Q7fi 
rial Moy i 7 / u 




550 


316 (250) 


Hanlev 1976 




1,380 


504 (103) 


Hanley 1976 




'150-330 (240) 




Pfister and others 1 977 


Pinus pondorosa 


'188 




(^larpv/ ^inH Athorc 1Q7^ 




'30-150 (188) 




Pfister and others 1 975 




490-570 


150-250 (150) 


Whittaker and Niering 1975 






50-250 (350+) 


Weaver and Forcella 1 977 


Festuca idahioensis 


235 


2.35 


Collins and Weaver 1978 




195 


1.95 


Weaver and Collins 1977 




147 


1.47 


Daubenmire 1970 


Festuca scabrella 


152 


1.52 


Willms and others 1986 


Bouteioua gracilis 


103 


1.03 


Weaver 1 983 




53-95 


0.5-1.0 


Hunt and others 1 988 


Atriplex spp. 


28 




West 1983 



'Since merchantable production and standing crop are emphasized in these studies, one might expect the figures to be 50 to 66 percent of those reported in 
studies of total production (Weaver and Forcella 1977). 

Where production and standing crop are expressed volumetrically, masses were calculated using specific gravities of ABLA=0.38, PIAL=0.40, PIEL=0.35, 
PSME=0.45, PIPO=0.43, PICO=0.38, THPL=0.33, TSHE=0.41 (U.S. FPL 1974). 



(starvation) may exclude, from an altitudinal zone, plants 
of lower zones. Average growing season temperatures in 
the alpine (and for most zones below it) are distinctly 
lower than temperatures in the vegetation zone below 
(fig. 3). This correlation suggests the possibility of — but 
does not prove — significant effects of growing season 
temperature. In support of this h3T)othesis, production- 
temperature relations (discussed below) suggest that the 
vegetation of each altitudinal zone is especially adapted 
to temperatures occurring in its zone. In this regard, it 
is satisfying to see that optimum temperatures for white- 
bark pine photosynthesis (20-25 °C, Jacobs and Weaver 
1990) are similar to summer maximum temperatures 
(10-25 °C, Weaver 1990). 

Graphs of estimated soil water availability against veg- 
etation type (fig. 4) indicate little or no stress in alpine, 
whitebark, subalpine fir, and Douglas-fir zones. On this 
basis one might exclude droiight as a factor determining 
whitebark's upslope limit. Due to two modifying factors, 



this conclusion seems premature. First, as one moves 
upslope from forest to alpine (or moimtain meadow), 
winds increase and, with increasing wind, drought in- 
creases due to reduced water availability — snow may be 
blown off site (Daubenmire 1981) and water in uninsu- 
lated soils freezes — and increasing water loss — through 
abrasion of cuticle (Hadley and Smith 1987) and thinning 
of the leafs boundary layer (Gates and Papian 1971; Nobel 
1983). I see winter wind effects as probable controllers be- 
cause trees invading (or planted into) mountain meadows 
are more often desiccated in winter than stunmer. The 
wind effect hypothesis is consistent with the fact that 
groups of trees sometimes invade sites that individuals 
cannot invade alone (Armand 1992; Tranquillini 1979). In 
such situations adjacent trees are more likely sheltering 
each other from drying wind than from lethal low tempera- 
tures. Second, as one moves upslope the average soil be- 
comes progressively better drained (Weaver 1979), and 
thus effective precipitation is a smaller fraction of total 



145 




10 20 30 



ECOSYSTEM TYPE 



Figure 1 — Midwinter low temperatures in 10 Rocky Mountain ecosystems. Low temperatures (°C 
and °F) are represented by absolute winter minimums (A, ABSMN) and average January minima 
(B, AVJMN). Ecosystem types range altitudinally from alpine (2), down through forests (8-20) to 
grass and shrublands (26-36); specifically, they are alpine (2), whitebark pine (8), subalpine fir (10), 
Douglas-fir (16), cedar-hemlock (18), ponderosa pine (20), Idaho fescue (26), bluebunch wheatgrass 
(28), gramagrass (30), and desert shrub (36). Large and small circles represent data gathered east 
and west of the Rockies, respectively. Lines connect median values. 




30 40' '0 10 

ECOSYSTEM TYPE 



Figure 2— The coldest early spring (A, SPR-FRST) and late fall (8, FALL-FRST) frosts (°C 
and "F) in 10 Rocky Mountain ecosystems. Ecosystems, symbols, and lines are as in figure 1. 




O 10 20 30 40 O 10 20 30 40 

ECOSYSTEM TYPE 

Figure 3 — Average growing season temperatures in 1 0 Rocky Mountain ecosystems. The first graph 
(A, TEMP gs) gives the average growing season temperature (°C and "F). Q gs in the second graph 
(B) considers the exponential response of physiological processes to increasing temperature (Q,q). 
Ecosystem types, symbols, and lines are as in figure 1 . 



146 



precipitation at higher than lower altitudes; that is, the 
Walter drought indices presented in figure 4 probably over- 
estimate water availability on (high-altitude) thin-soil sites. 

Downward Limits 

I speculate that downward limits of whitebark pine 
might be determined by heat, drought, or competition. 
The following paragraphs consider these hypotheses. 

As one moves downslope maximum temperatures 
(fig. 5) rise. That this correlate of tree disappearance is 
not controlling is suggested by three facts. Summer aver- 
ages in vegetation zones below whitebark pine (10-15 °C, 



fig. 3) are well below the whitebark pine optimimi 
(20-25 °C, Jacobs and Weaver 1990), so downward migra- 
tion might actually improve production. July maxima 
(20-27 °C, fig. 5) in zones below are near the whitebark 
pine optimvuu temperature (20-25 °C, Jacobs and Weaver 
1990) and thus should not be damaging. Long-term 
maxima (30-42 °C, fig. 5) in zones below do not eliminate 
net photosynthesis (Jacobs and Weaver 1990) and thus 
are probably not lethal. In addition, if whitebark pine 
trees are well watered, they grow well as lawn trees in the 
Agropyron spicatum zone (for example, Belgrade, MT) 
where temperatures are far higher than those fovmd on 
sites whitebark naturally occupies. 




ECOSYSTEM TYPE 

Figure 4 — Drought severity in 10 Rocky Mountain ecosystems. The first graph (A) 
gives the number of dry months (DRT-MNTHS). The second (B) indexes the an- 
nual water deficiency (W-DFCT, mm). Both were calculated after Walter (1973). 
The ecosystem types, symbols, and lines are as in figure 1 . 




0 10 20 30 40 0 10 20 30 40 

ECOSYSTEM TYPE 

Figure 5 — High temperatures (°C and °F) recorded in 10 Rocky Mountain ecosys- 
tems. The first graph (A) represents absolute summer highs (ABSMX). The sec- 
ond (B) gives average July maxima (AVJMX). The ecosystem types, symbols, and 
lines are as in figure 1 . 



147 





ECOSYSTEM TYPE 



Figure 6— Biomass production in 10 Rocky Mountain ecosystems (see also table 2). The 
first graph (A) presents harvestable production (HPRD, g/m^/yr). The second (B) gives to- 
tal production (TPRD, g/m^/yr). The ecosystem types, symbols, and lines are as in 
figure 1 . 



Drought likely sets the lower physiological limit 
(Hutchinson 1958) for whitebark pine's downward exten- 
sion. Walter's (1973) index suggests the drought months 
and water deficits are near zero in high forests (whitebark 
pine, subalpine fir, and Douglas-fir), greater in ponderosa 
pine forests and grasslands, and still greater in Atriplex 
shrublands (fig. 4). If (as noted above) wind decreases 
and waterholding capacity of soils increases downslope, 
water conditions for whitebark production improve as one 
moves downslope; this is consistent with the fact that 
whitebarks growing in the subalpine fir zone can be stately 
timber producers (Pfister and others 1977; Weaver and 
Dale 1977). Water deficits remain slight down through the 
subalpine fir and possibly the Douglas-fir zones. The com- 
plete absence of whitebarks in the ponderosa pine and 
grassland zones, despite possible nutcracker dispersal 
(Lanner 1990; Tomback and others 1990), supports my 
doubt that the tree is capable of growing in the ponderosa 
pine and drier grassland-desert zones. Similarly, in the 
geographic dimension, the southern and eastern limits of 
whitebark pine seem to appear where precipitation in its 
altitudinal zone becomes too low (Arno and Weaver 1990). 

The lower realized limit (Hutchinson 1958) of white- 
bark is set by competition with subalpine fir and lodge- 
pole pine rather than drought. In our region, evidence for 
this fact lies in the tree's existence and good growth on 
sites in the subalpine fir zone (but apparently not in the 
Douglas-fir zone) where subalpine fir and lodgepole com- 
petition have been removed by clearcutting or fire. The 
lower limit of whitebark is, then, set by those factors — 
likely drought (thin soils and wind) and low summer tem- 
perature — that exclude subalpine fir and lodgepole pine 
from higher sites. In the geographic dimension, to the 
north and west where rainfall is higher, whitebark is also 
unable to form pure stands or disappears (Arno and 
Weaver 1990). (European foresters ask why we don't 
eliminate competitive "natives" [subalpine fir and lodgepole 
pine] from their zone to allow expansion of whitebark pine 
and increased production of quality white pine lumber.) 



Production 

As one moves from dry grasslands to moist forests and 
upward to alpine sites one sees production rise and fall 
again. Harvestable production — merchantable standing 
crop divided by years in its production — rises from dry 
grasslands (100 g/mVyr) to moist forests (300 g/mVyr) and 
falls through high forests to the alpine (100 g/mVyr, fig. 6, 
table 2). Available data suggest that total aboveground 
production (total increment ^ro?lyr) also increases fi-om 
dry grass and shrublands (50-100 g/mVjr) to warm forests 
(over 500 g/mVyr) and falls to the alpine (100 g/mV5T, 
fig. 6, table 1). The trend is parallel, but — at least in for- 
ests — total production is far higher than harvestable pro- 
duction. Three factors may contribute to this difference: 
(1) Much of the production, over the life of a stand, is lost 
to needle drop, self pruning, and natural thinning; this 
material decomposes and is recycled. The total produc- 
tion rates reported are deceptively high because they in- 
clude nutrients being recycled to the atmosphere and soil; 
that is, this production consists of an actual production 
component based on nutrient import (weathering and at- 
mospheric delivery) and a restructuring component based 
on recycling of nutrients from "obsolete" leaves, branches, 
and trees. (2) Only 33 to 50 percent of the material extant 
at harvest is merchantable (Weaver and Forcella 1977). 
(3) Most of the total production data comes from areas 
outside our region. If the climates of Douglas-fir, cedar- 
hemlock, and ponderosa pine stands from which produc- 
tion data are reported are moister or warmer than the 
Montana stands in which the harvestable productivity 
data were gathered, somewhat higher productivities 
might be expected. 

I expect productivity to be controlled, like vegetation 
distribution, by climate. In the following paragraphs I 
will explore the h3rpothesis that production depends on 
growing season length and the warmth of that season. 

Growing season length might be indexed as the number 
of months where daily average temperatures are above 



148 



Table 3 — Relationship of production (g/m^/yr) to climatic factors 



Harvestable production' 



HP 


66.00 


+ 


12.8000 


WS 


p = 0.50 


= 0.05 




-171.00 


+ 


70.0000 


GS 


p = 0.00 


= 0.86 




34.00 


+ 


1.7700 


GSxT 


p = 0.04 


/■2 = 0.44 




-38.00 


+ 


14.4000 


GSxQ 


p = 0.01 


= 0.58 


Log HP = 


0.78 






wo 


n - n nn 


f2 _ n 7fi 




1 75 




n nnfin 




n — n 1 5 


f2 — n 97 




1.44 


+ 


0.0530 


GSxQ 


p = 0.03 


r2 = 0.44 


tal Droduction^ 














TP 


-880.00 


+ 


321.0000 


GS 


p = 0.16 


r2 = 0.22 




50.00 


+ 


8.3200 


GSxT 


p = 0.32 


= 0.12 




-398.00 


+ 


77.0000 


GSxQ 


p = 0.19 


r2 = 0.20 


Log TP = 


0.40 


+ 


0.4700 


GS 


p = 0.03 


= 0.45 




1.96 


+ 


0.0086 


GSxT 


p = 0.32 


r2 = 0.12 




1.47 


+ 


0.0820 


GSxQ 


p = 0.22 


^ = 0.22 



'Harvestable production (HP, lop diameter greater than 10 cm) and the log of harvestable production 
(log HP) are regressed against warm season (WS), warm-moist growing season (GS), the product of 
warm-moist growing season and average growing season temperature (GSxT), and the product of 
wann-moist growing season and Q,o-based growth support units (GSxQ). 

n"otal production (TP) and log total production (log TP) are also regressed against GS, GSxT, and 
GSxQ. 



0 °C. The rationale is that, iinless air temperatiires are 
above 0 °C, exposed soils will be frozen (Weaver, in prepa- 
ration) and frozen soils will deliver little water and nutri- 
ents. Because a regression of production against this in- 
dex of growing season is so loose (r^ = 0.05, table 3, fig. 7), 
the hypothesis that warm-season length controls produc- 
tion is inadequate. 

The previous index may fail because, at least at lower 
altitudes, growing season is limited in late summer by a 
lack of soil water, rather than temperatiire. If so, the 
growing season index would be improved as an index of 
production by subtracting soil-dry months from total soil- 
warm months. As above, I index drought as periods when 
precipitation (mmy2) is less than average temperature (C, 
after Walter 1973). Regressions of production against 
months without either temperature or moisture stress ex- 
plain most of the variance in harvestable production (r^ = 
0.86, table 3, fig. 7). Season length is a weaker correlate 
of total production (r^ = 0.22). I offer two unsatifying 
hypotheses for this important difference: (1) Growing sea- 
son predicts harvestable production better than total pro- 
duction because photosynthates are allocated first to 
canopy maintainence (so this component does not vary 
much among types) and second to supporting structure; if 
so, the relationship between season length and tnmk pro- 
duction is clarified by omitting the common leaf-twig pro- 
duction. (2) If total production were measured in slightly 
more favorable segments of the environmental types than 
harvestable production, the iinderstatement of season 
length would weaken the relationship. 

If all vegetation — grassland to forest to alpine — had a 
single temperature response curve, we would expect pro- 
duction in "warm moist soil days" to rise exponentially 
with increasing temperatiu-e. This expectation is based 



on the rise in chemical reaction rates with rising tem- 
perature, on the resultant physiological response curves 
(Larcher 1975), on degree-day prediction of plant phenol- 
ogy (Chang 1960), and on the prediction that degree-day 
formulae might be improved by use of temperatiire aver- 
ages computed from a nonlinear Q^^ relationship (Larcher 
1975). Thus, we expect production to be better correlated 
with the product of growing season days x temperature 
than with growing season days alone. In fact, whether 
we use our average growing season temperature or grow- 
ing season Q index, the relationship is poorer for both 
harvestable (r^= 0.44) and total (r^= 0.12) production. 
Since the basic physiological response is a physiochemical 
necessity, I deduce first that, while the vegetation of each 
environmental zone does respond to temperature, it is 
specifically adapted to temperattires in that zone and, sec- 
ond, that temperature adaptation distinct to each zone 
eliminates temperature fi'om the predictive equation. I 
doubt that temperature would be eliminated if adaptation 
were eliminated by using a genetically homogeneous veg- 
etation type; for example, if the region were vegetated 
with a single crop, production would be best predicted by 
the product of growing season length and temperature. 
Since competitiveness of a tree undoubtedly depends on 
its productivity relative to competitors, this conclusion 
supports earlier speculation that growing season tempera- 
ture adaptation provides at least one basis for the 
existance of different vegetation zones. 

While I have argued that production in a zone is best 
predicted from the length of its warm moist growing sea- 
son, I expect the standing crop of mature accvmaulative 
(woody) vegetation (table 2) to be largely determined by 
site fertility. Accmnulation — slow or fast — depends on 
photosynthesis minus respiration. Synthesis depends on 



149 



light, warmth, water (open stomates), and nutrients. In 
the absence of stand destruction, annual inputs of light, 
warmth, water, and elements with atmospheric cycles 
(for example, C, H, O, N) should support eternal accumu- 
lation. Limited supplies of nutrients with geologic cycles 
(for example, P, K, Ca) will halt accumulation (Weaver 
1978). It is also argued (Odum 1969) that accvunulation 
is halted when respiration equals production. Thus, in 
young trees net production is high because the ratio of 
photosynthetic mass to respiratory mass is high, in longer 
stemmed trees efflcency declines as the respiratory load 
increases, and growth ceases when respiration equals 
gross production; heartwood's contribution to this rela- 
tionship is in proportion to its inertness. While the respi- 
ratory factor undoubtedly contributes to observed declines 
in production with stand height (and age) (Weaver and 
others 1990), I see nutrient supplies as more limiting in 
high forests because the trees are far shorter (and there- 
fore more efficient) than productive trees of lower alti- 
tudes. In contrast, standing crops in less accumulative 
(herbaceous) vegetation are determined by production oc- 
curring in one growing season, and while it could be de- 
termined by supplies of a geologically cycling nutrient, 
standing crop is more likely determined by a bulk re- 
source like light (unlikely), warmth, water, or an atmo- 
spheric nutrient like nitrogen. 



CONCLUSIONS 

Whitebark pine vegetation is likely excluded from the al- 
pine zone by cool growing season temperatiires or droughts 
occurring most likely in winter. It seems unlikely that 
freezes of late fall, midwinter, or early spring exclude the 
tree from higher sites. 

The lower physiological limit of whitebark pine is likely 
set by drought. Its lower realized limit is likely due to com- 
petition with lodgepole pine or subalpine fir. By control of 
competition, managers could probably extend whitebark's 
range downslope. 

The distribution of other dominant plants on the tem- 
perature water gradient may be similarly controlled. 

Annual production is strongly related (H = 0.86) to the 
number of growing season days. The lack of correlation 
with temperature suggests that plants of any vegetation 
zone are adapted to temperature conditions peculiar to 
that zone. Mature standing crops of woody vegetation are 
more likely determined by nonatmospheric nutrients than 
temperature and precipitation. Potential standing crops 
of herbaceous vegetation, on the other hand, are more 
likely limited by bulk resources such as warmth, water, 
carbon, or nitrogen. 




3 6 

WM SEASON, mo 




20 40 

GS X T (moxC 5) 



\ 

o 

cr 

CL 
X 



r2= 0.85 



B 



J L 




J L 



J L 



3 6 

GRO SEASON, mo 




GS X Q (moxQ) 



Figure 7 — The relationship of harvestable production (HPROD, g/m^/yr) to season 
length and temperature: (A) production vs. warm season (WM-SEASON), (B) produc- 
tion vs. warm-moist season (GRO-SEASON), (C) production vs. warm-moist season 
(GS) X average summer temperature above 5 °C (C-5), (D) production vs. warm moist 
season (GS) x growth support units (Qg^). The ecosystem types are as in figure 1 . 



150 



ACKNOWLEDGMENTS 

I thank colleagues who helped me locate weather sta- 
tions representative of major environmental types of the re- 
gion (S. Amo, D. Despain, J. Habeck, R. Rankin, and others 
acknowledged in Weaver 1980 and 1990), acquire the data 
(J. Caprio), and choose parameters for study (J. Pickett and 
J. Brown). The critical review of P. Burgeron, D. Despain, 
S. Harvey, W. Moir, and D. Perry is greatly appreciated. I 
am also grateful for the support provided by Intermoimtain 
Research Station, USDA Forest Service (INT-92720- 
RJVA). 

REFERENCES 

Aplet, G.; Smith, F.; Laven, R. 1989. Stemwood biomass 
and production during spruce-fir stand development. 
Journal of Ecology. 77: 70-77. 

Armand, A. 1992. Sharp and gradual motmtain timber- 
lines as a result of species interaction. In: Hansen, A.; 
DiCastri, F. 1992. Landscape boundaries. New York: 
Springer Verlag: 360-378. 

Amo, S. 1984. Timber line — moimtain and arctic forest 
frontiers. Seattle: Moimtaineers. 304 p. 

Amo, S.; Weaver, T. 1990. Whitebark pine commimity 
types and their pattem in the landscape. In: Schmidt, 
W.; McDonald, K, comps. Proceedings — symposium on 
whitebark pine ecosystems: ecology and management of 
a high-moimtain resource. Gen. Tech. Rep. INT-270. 
Ogden, UT: U.S. Department of Agriculture, Forest 
Service, Intermovmtain Research Station: 97-105. 

Burke, M.; Gusta, L.; Quamme, H.; Weiser, C.; Li, P. 
1976. Freezing injxiry in plants. Annual Review of Plant 
Physiology. 27: 507-528. 

Chang, J. 1968. Climate and agriculture. Chicago: Aldine. 
304 p. 

Collins, D.; Weaver, T. 1978. Effects of summer weather 
modification (irrigation in Festuca idahoensis-Agro- 
pyron spicatum grasslands. Jovu-nal of Range Manage- 
ment. 31: 264-269. 

Daubenmire, R. 1943. Vegetation zonation in the Rocky 
Mountains. Botanical Review. 9: 325-393. 

Daubenmire, R. 1956. Climate as a determinant of vegeta- 
tion distibution in eastern Washington and northern 
Idaho. Ecological Monographs. 26: 131-154. 

Daubenmire, R. 1970. Steppe vegetation of Washington. 
Bull. 62. Pullman, WA: Washington Agricultural Ex- 
periment Station. 131 p. 

Daubenmire, R. 1981. Subalpine parks associated with 
snow transfer in the mountains of northern Idaho and 
eastem Washington. Northwest Science. 55: 124-135. 

Daubenmire, R.; Daubenmire, J. 1968. Forest vegetation 
of eastern Washington and northern Idaho. Bull. 60. 
Pullman, WA: Washington Agricultural Experiment 
Station. 104 p. 

Forcella, F.; Weaver, T. 1977. Biomass and productivity of 
the subalpine Pinus albicaulis/Vaccinium scoparium as- 
sociation, MT USA. Vegetatio. 35: 95-105. 

Gams, H. 1931. Die klimatische Begrenzimg von Pflanzen 
arealen und die Verteilvmg der hygrischen Konti- 
nentalitat in den Alpen. Zietschrit der Gesellschafl fur 
Erdkunde. 9: 32-68. 



Gates, D.; Papian, L. 1971. Atlas of energy budgets of 
plant leaves. London and New York: Academic Press. 
277 p. 

Geiger, R. 1965. The climate near the ground. Cambridge, 

MA: Harvard University Press. 611 p. 
Hadley, J.; Smith, W. 1987. Influence of knunmholz 

microclimate on needle physiology and survival. 

Oecologia. 73: 82-90. 
Hanley, D. 1976. Tree biomass and production estimates 

for three habitat types in northern Idaho. Bull. 14. 

Moscow, ID: University of Idaho, College of Forestry. 

16 p. 

Hunt, W.; Ingham, E.; Coleman, D.; ElHott, E.; Reid, P. 
1988. Nitrogen limitation of production and decomposi- 
tion in prairie, moimtain meadow, and pine forest. 
Ecology. 69: 1009-1016. 

Holdridge, L. 1967. Determination of world plant forma- 
tions from simple climatic data. Science. 130: 572. 

Huschle, G.; Hironaka, M. 1980. Classification and ordi- 
nation of serai plant communities. Journal of Range 
Management. 33: 179-182. 

Hutchinson, G. 1958. Concluding remarks. Cold Springs 
Harbor Symposium in Quantiative Biology. 22: 415-427. 

Jacobs, J.; Weaver, T. 1990. Effect of temperature and 
temperature preconditioning on seedlimg performance 
of whitebark pine. In: Schmidt, W.; McDonald, K, 
comps. Proceedings — symposium on whitebark pine eco- 
systems: ecology and management of a high-mountain 
resource. Gen. Tech. Rep. INT-270. Ogden, UT: U.S. 
Department of Agriculture, Forest Service, Intermoun- 
tain Research Station: 134-139. 

Kuchler, A. 1964. Potential vegetation of the contermi- 
nous US. Spec. Publ. 36. New York: American Geo- 
graphical Society. 155 p. 

Landis, T.; Mogren, E. 1975. Tree strata biomass of sub- 
alpine spmce-fir stands in SW Colorado. Forest Science. 
21: 9-12. 

Lanner, R. 1980. Biology, taxonomy, evolution, and geog- 
raphy of the stone pines of the world. In: Schmidt, W.; 
McDonald, K., comps. Proceedings — symposium on 
whitebark pine ecosystems: ecology and management of 
a high-mountain resource. Gen. Tech. Rep. INT-270. 
Ogden, UT: U.S. Department of Agriculture, Forest Ser- 
vice, Intermountain Research Station: 14-24. 

Larcher, W. 1975. Physiological plant ecology. New York: 
Springer. 252 p. 

Neilson, R. 1986. High resolution climatic analysis and 
southwest biogeography. Science. 232: 27-33. 

Nielson, R. 1992. Toward a rule based biome model. Land- 
scape Ecology. 7: 27-43. 

Nobel, P. 1983. Biophysical plant physiology. San 
Francisco: Freeman. 608 p. 

Penman, H. 1949. The dependence of transpiration on 
weather and soil conditions. Journal of Soil Science. 
1: 74-89. 

Pfister, R.; Kovalchik, B.; Amo, S.; Presby, R. 1977. For- 
est habitat types of Montana. Gen. Tech. Rep. INT-34. 
Ogden, UT: U.S. Department of Agriculture, Forest Ser- 
vice, Intermountain Forest and Range Experiment Sta- 
tion. 174 p. 



151 



Rodin, L.; Bazilevich, N. 1967. Production and mineral cy- 
cling in terrestrial vegetation. London: Oliver and Boyd. 
288 p. 

Salisbury, F.; Ross, C. 1992. Plant physiology. Belmont, 
CA: Wadsworth. 682 p. 

Scott, D.; Billings, W. 1964. Effects of environmental fac- 
tors on standing crop and production of an alpine tun- 
dra. Ecological Monographs. 34: 243-270. 

Stephenson, N. 1990. Climatic control of vegetation distri- 
bution: the role of water balance. American Naturalist. 
135: 649-670. 

Thilenius, J.; Smith, D.; Brown, G. 1974. Effect of 2,4-D 
on composition and production of an alpine plant com- 
munity in Wyoming. Journal of Range Management. 
27: 140-142. 

Thornthwaite, C. 1948. An approach toward a rational 
classification of climate. Geographical Review. 38: 
55-94. 

Tomback, D.; Hoffmann, L.; Sund, S. 1990. Coevolution of 
whitebark pine and nutcrackers: implications for forest 
regeneration. In: Schmidt, W.; McDonald, K., comps. 
Proceedings — sjrmposium on whitebark pine ecosys- 
tems: ecology and management of a high-mountain re- 
source. Gen. Tech. Rep. INT-270. Ogden, UT: U.S. De- 
partment of Agriculture, Forest Service, Intermountain 
Research Station: 118-129. 

Tranquillini, W. 1979. Physiological ecology of the alpine 
timberline. Tree existence at high altitude with special 
reference to the European Alps. New York: Springer. 
137 p. 

U.S. FPL. 1974. Wood handbook: wood as an engineering 
material. Agric. Handb. 72. Washington, DC: U.S. De- 
partment of Agriculture, Forest Service, Forest Prod- 
ucts Laboratory, [n.p.]. 

USGS. 1970. The national atlas. Washington, DC: U.S. 
Department of the Interior, Geological Survey. 417 p. 

Walter, H. 1973. Vegetation of the earth in relation to cli- 
mate and ecophysiological conditions. New York: 
Springer. 237 p. 

Weaver, T. 1978. Changes in soils along a vegetation- 
altitudinal gradient of the Northern Rocky Mountains. 
In: Youngberg, C, ed. Forest soils and land use. Pro- 
ceedings; 5th North American Soils Conference. Fort 
Collins, CO: Colorado State University, Forestry 
Department: 14-29. 



Weaver, T. 1980. Climates of vegetation types of the 
northern Rocky Mountains and adjacent plains. Ameri- 
can Midland Naturalist. 103: 392-398. 

Weaver, T. 1983. Yield response of high plains grasslands 
to water enrichment, three phases. In: Holman, L.; 
Knutsen, G. State of Montana activities in the High 
Plains Cooperative Program 1981-1983. Helena, MT: 
Montana Department of Natural Resources, Water 
Resources Division: 77-94. 

Weaver, T. 1990. Climates of subalpine pine woodlands. 
In: Schmidt, W.; McDonald, K., comps. Proceedings — 
S3mtiposium on whitebark pine ecosystems: ecology and 
management of a high-mountain resource. Gen. Tech. 
Rep. INT-270. Ogden, UT: U.S. Department of Agricul- 
ture, Forest Service, Intermountain Research Station: 
72-79. 

Weaver, T.; Collins, D. 1977. Effects of weather modifica- 
tion (increased snow pack) on Festuca idahoensis mead- 
ows. Journal of Range Management. 30: 451-456. 

Weaver, T.; Dale, D. 1974. Pinus albicaulis in central 
Montana: environment, vegetation, and production. 
American Midland Naturalist. 92: 222-229. 

Weaver, T.; Forcella, F. 1977. Biomass of fifty conifer for- 
ests and nutrient exports associated with their harvest. 
Great Basin Naturahst. 37: 395-401. 

Weaver, T.; Forcella, F.; Dale, D. 1990. Stand develop- 
ment in whitebark pine woodlands. In: Schmidt, W.; 
McDonald, K., comps. Proceedings — symposium on 
whitebark pine ecosystems: ecology and management 
of a high-mountain resource. Gen. Tech. Rep. INT-270. 
Ogden, UT: U.S. Department of Agriculture, Forest Ser- 
vice, Intermountain Research Station: 151-155. 

West, N. 1983. Temperate deserts and semideserts: Inter- 
mountain salt desert shrubland. Amsterdam: Elsevier: 
382. 

Whittaker, R. 1975. Communities and ecosystems. New 

York: Macmillan. 385 p. 
Whittaker, R.; Niering, W. 1975. Vegetation of the Santa 

Catalina Mountains, Arizona. V. Biomass, production 

and diversity along the elevation gradient. Ecology. 56: 

771-790. 

Willms, W.; Smoliak, S.; Bailey, A. 1986. Herbage produc- 
tion following litter removal on Alberta native grass- 
lands. Journal of Range Management. 39: 536-540. 



152 



Regeneration 




153 



A 



VARIATION IN SIZE AND WEIGHT OF 
CONES AND SEEDS IN FOUR NATURAL 
POPULATIONS OF CARPATHIAN 
STONE PINE 

I. Blada 
N. Popescu 



Abstract — High variation was found within populations of 
stone pine (Pinus cembra L.) for seeds per cone and seeds per 
cone weight, but middle high for 1,000-seed weight and low for 
cone length and cone diameter. The cone length, cone diameter, 
seeds per cone, seeds per cone weight, and 1,000-seed weight 
means were 4.7 cm, 4.2 cm, 37.8 cm, 9.2 g, and 250 g, respec- 
tively. Southern populations were similar in cone length and 
cone diameter, but not in seeds per cone, seeds per cone weight, 
and l.OOO-seed weight. Cone length and cone diameter from 
open-pollinated, cross-pollinated and self-pollinated cones were 
similar, while seeds per cone, seeds per cone weight, and 1,000- 
seed weight, were not. All traits displayed continuous variation. 
Significant correlations were found among all traits, except 
1,000-seed weight. 



Stone pine (Pinus cembra L.) is naturally distributed 
in the highest forest zone of the Alps and Carpathian 
Mountains (Critchfield and Little 1966; Holzer 1975). 
In the Alps, the low-elevation stands range between 1,100 
to 1,500 m, but the main zone extends between 1,700 to 
2,000 m (Contini and Lavarelo 1982; Holzer 1975), while 
the high-elevation form of the species climbs as single 
trees up to 2,700 m above sea level (Moser 1960). In 
Romania, stone pine ranges from 1,350 to 1,880 m in the 
northern Carpathians (Gubesch 1971) and from 1,350 to 
1,980 m in the southern Carpathians (Beldie 1941; Oarcea 
1966; Tataranu and Costea 1952). 

Because of its tolerance for low temperatvires, the spe- 
cies is very important for reforestation of the subalpine 
zone; in this zone it is also important on watersheds, for 
stabilizing avalanche areas, and for reducing the effects 
of flash floods (Holzer 1972). 

Stone pine has a particular importance for the silvicul- 
ture of the subalpine zone of the Carpathians. For this 
reason a genetic improvement program with both intra- 
and interspecific crosses is being carried out (Blada 1982, 
1990a); some results have been published (Blada 1987, 
1990b, 1992a) or are in preparation (Blada 1992b). 

This paper reports on the phenotypic variation in size 
and weight of cones and seeds in four natural populations 
of Carpathian stone pine, as part of the program. 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

I. Blada and N. Popescu are Forest Geneticists, Forest Research Insti- 
tute, Bucharest 11, Romania. 



MATERIALS AND METHODS 

Twenty trees were sampled at random in each of the 
four populations listed in table 1. In mid-July 1991, 25 
cones from open pollination (OP) on each of the 20 trees 
(total 80 trees) were protected against the mountain jay 
(Nucifraga caryocatactes L.) by using metal net bags 25 
by 20 cm in size. 

For comparative reasons, cones and seeds obtained from 
controlled cross- and self-pollinations (CP, SP) in the 
Gemenele population were measured; these pollinations 
were performed in July 1989 using 10 x 10 full diallel 
mating design (Blada 1992b), according to Griffing's 
(1956) Experimental Method 1. 

Cone and seed measurements were taken soon afler 
their collection in October 1990 for controlled cross- 
pollinated and self-pollinated and in October 1991 for 
open-pollinated cones. Twenty-five cones were measured 
from each tree resulting fi*om open-pollinated cones, and 
20 cones fi-om each combination resulting fi*om controlled 
cross-pollinated and self-pollinated cones. Five traits 
were measured as shown in table 2. 

Using data from measurements, the population mean 
(x), standard deviation (g), mean square {&), variation co- 
efficient (VC), range of variation (Q), and correlation coef- 
ficient (r) were calculated. 

WITHES-POPULATION VARIATION 

Statistical parameters of the cones and seeds of the four 
Carpathian stone pine populations listed in table 1 are 
summarized in table 3 and figure 1. 



Table 1 — Geographic distribution of Pinus cembra studied 
populations in the Carpathians 



Population 


Latitude N. Longitude E. 


Altitude 


Carpathian 
Range 




Degrees 


Meters 




Gemenele 


45°35" 22°50' 


1,720 


Southern 


Stana de Rau 


45°25' 23°03' 


1,450 


Southern 


Pietrele 


45°23' 22°52' 


1,550 


Southern 


Lala 


47°33' 25°05' 


1,520 


Northern 



154 



Table 2 — Measured traits 



Trait Units Symbols 



Cone length Centimeters CL 

Cone diameter Centimeters CD 

Seeds per cone Number SC 

Seeds per cone weight Grams SCW 

1 ,000-seed weight Grams 1 ,000-SW 



The following main results were obtained for the 
Gemenele, Stana de Rau, Pietrele, and Lala populations 
(in sequence): 

• The cone length mean was 4.9 ± 0.4 cm, 4.7 ± 0.4 cm, 
4.9 ± 0.6 cm, and 4.1 ± 0.4 cm, respectively. 

• The cone diameter mean was 4.2 ± 0.3 cm, 4.1 ± 0.3 cm, 
4.1 ± 0.3 cm, and 3.5 ± 0.2 cm, respectively. 

• The seeds per cone mean was 40.9 ± 9.6, 36.7 ± 8.1, 
52.6 ± 13.3, and 21.1 ± 7.1, respectively. 

• The seeds per cone weight was 9.8 ± 2.7 g, 9.0 ± 2.5 g, 
12.5 ± 3.9 g, and 5.6 ± 1.7 g, respectively. 



Table 3 — Mean (x), phenotypic standard deviation (a), mean square 
(cf), variation coefficient {VQ, range of variation (O) 



Parameters (open pollination) 



Trait 


x± a 




VC 




Q 








Gemenele population 








CL 


4.9 


± 0.4 


0.15 


7.9 


4.1 




5.8 


CD 


4.2 


+ 0.3 


0.11 


7.9 


3.6 




5.0 


SC 


40.9 


± 9.6 


91.31 


23.3 


22.5 




60.2 


SCW 


9.8 


± 2.7 


7.50 


28.0 


4.6 




18.1 


1,000-SW 


238.0 


± 47.4 


2,243 


19.9 


170 




352 






Stana de Rau population 








CL 


4.7 


± 0.4 


0.18 


9.1 


3.8 




5.6 


CD 


4.1 


± 0.3 


0.08 


7.0 


3.7 




4.9 


SC 


36.7 


± 8.1 


65.67 


22.1 


21.7 




55.4 


SCW 


9.0 


± 2.5 


6.05 


27.3 


5.5 




14.1 


1,000-SW 


252 


± 37.6 


1,412 


14.9 


175 




293 






Pietrele population 










CL 


4.9 


± 0.6 


0.34 


11.7 


4.1 




6.0 


CD 


4.1 


± 0.3 


0.10 


7.7 


3.4 




4.7 


SC 


52.6 


± 13.3 


177.31 


25.3 


31.2 




90.4 


SCW 


12.5 


± 3.9 


15.27 


31.3 


6.2 




23.1 


1 ,000-SW 


238 


± 60.9 


3,712 


25.6 


138 




428 








Lala population 










CL 


4.1 


± 0.4 


0.17 


9.9 


3.7 




5.1 


CD 


3.5 


± 0.2 


0.04 


6.0 


3.1 




3.9 


SC 


21.1 


± 7.1 


51.15 


34.0 


13.6 




34.8 


SCW 


5.6 


± 1.7 


2.78 


29.9 


3.6 




9.4 


1 ,000-SW 


270 


± 46.5 


2,163 


17.2 


184 




346 




Trait parameters across four populations 






CL 


4.7 


± 0.5 


0.30 


11.7 


3.7 




6.0 


CD 


4.2 


± 0.3 


0.11 


7.9 


3.1 




5.0 


SC 


37.8 


± 14.9 


222.22 


39.4 


13.6 




90.4 


SCW 


9.2 


± 3.7 


13.77 


40.3 


3.6 




23.1 


1 ,000-SW 


250 


± 49.7 


2,471.00 


19.8 


167.0 




354.0 



• The 1,000-seed weight mean was 238 + 47.4 g, 252 ± 
37.6 g, 238 ± 60.9 g, and 270 ± 46.5 g, respectively. 

Therefore, the southern populations were similar in 
cone length and cone diameter, but not in seeds per cone, 
seeds per cone weight, and 1,000-seed weight. On the 
other hand, the southern populations ranked high and the 
northern population ranked low for all measured traits. 

Very high variation coefficients were found within 
each population for both seeds per cone and seeds per 
cone weight, but middle high for 1,000-seed weight and 
low for cone length and cone diameter traits. 

The last three columns of table 3 give information con- 
cerning mean squares, variation coefficients, and the 
range of variation for pop\ilation traits. 

ACROSS-POPULATION 
PARAMETERS 

The mean values and ranges of the traits across foiu* 
studied populations — and by extrapolation — for all the 
Carpathian stone pine were as follows (table 3, lower 
part): 

• The cone length mean ranged from 3.7 to 6.0 cm, with 
a mean of 4.7 ± 0.5 cm, and the cone diameter ranged 
from 3.1 to 5.0 cm, with a mean of 4.2 ± 0.3 cm. According 
to Contini and Lavarelo (1992), the size of cones from the 
Alps ranged from 5.0 to 10.0 cm in length and from 4.0 to 
6.0 cm in diameter. Therefore, the cone size from the Alps 
ranked high and the cones from the Carpathians ranked 
low. 

• The seeds per cone ranged from 13.6 to 90.4, with a 
mean of 37.8 ± 14.9. However, the seeds per cone mean 
from the Alps varied between 46 and 164 with a mean 
of 93 seeds (Rohmeder and Rohmeder 1955). Thus, the 
seeds per cone from the Alps ranked high and the seeds 
per cone from the Carpathians ranked low. 

• The seeds per cone weight mean ranged from 3.6 to 
23.1 g, with a mean of 9.2 ± 3.7 g. 

• The 1,000-seed weight mean ranged from 167 to 
354 g, with a mean of 250 ± 49.7 g. Consequently, a kilo- 
gram of seed from the Carpathians could include from 
2,825 to 5,988 seeds, with a mean of 4,000 seeds. Accord- 
ing to Rohmeder and Rohmeder (1955), the 1,000-seed 
weight mean from the Alps ranged between 150 and 
350 g; thus, the Carpathians 1,000-seed weight mean 
was very close to that of the Alps. 

The cone length, cone diameter, seeds per cone, seeds 
per cone weight, and 1,000-seed weight coefficients ac- 
counted for 11.7 percent, 7.9 percent, 39.4 percent, 40.3 
percent, and 19.8 percent of the variation, respectively 
(table 3, lower part). 

POLLINATION COMPARISONS 

The cone and seed parameters shown in table 4 came 
from records of 2,000 cones from 80 open-pollinated (OP) 
trees, 1,800 cones from a 10 x 10 full diallel mating design 
for cross-pollination (CP), and 200 cones from 10 self- 
pollinated (SP) trees, all from the Gemenele population. 
For example, the cone length ranged from: 



155 



3- 



2. 



Oi 

c 




300, 



250. 



200. 



.C150. 

Jioo 



o 
o 

° 50 



I 

/ 



I 

s 

; 

; 

s 
s 
s 
s 



1 2 3 4 X X X 



1 2 3 A X X X 



2 3 A X X X 



1 2 3 A 5f X f 



Figure 1 — Mean performance of five measured traits of cones and seeds from open pollination_ 
in four populations (1 , 2, 3, 4) compared to controlled cross-pollination (f ) and self-pollination (^) 
(x= average across 1 , 2, 3, 4 populations). 



4.1 to 5.8 cm, with a mean of 4.9 ± 0.4 cm for open- 
pollinated; 

3.5 to 6.0 cm, with a mean of 4.8 ± 0.6 cm for cross- 
pollinated; 

3.7 to 5.5 cm, with a mean of 4.7 ± 0.6 cm for self- 
pollinated. 

Variation coefficients (VC) for open-pollinated, cross- 
pollinated, and self-pollinated were 7.9 percent, 12.8 per- 
cent, and 12.1 percent, respectively. 



Table 4 — Comparison between statistical parameters of cones and seeds in 
the Gemenele population, according to the pollination type 



Type of Parameters 



Trait 


pollination^ 


k±a 




VC 




0 




CL 


OP 


4.9 


± 


0.4 


0.15 


7.9 


4.1 




5.8 




CP 


4.8 


± 


0.6 


0.37 


12.8 


3.5 




6.0 




SP 


4.7 


± 


0.6 


0.36 


12.1 


3.7 




5.5 


CD 


OP 


4.2 


+ 


0.3 


0.11 


7.9 


3.6 




5.0 




CP 


4.3 


+ 


0.4 


0.15 


9.1 


3.7 




5.4 




SP 


4.2 


± 


0.4 


0.16 


9.4 


3.6 




4.9 


SC 


OP 


40.9 


+ 


9.6 


91.31 


23.3 


22.5 




60.2 




CP 


62.0 


± 


16.5 


272.24 


26.6 


41.4 




96.4 




SP 


64.0 


+ 


12.9 


167.12 


20.3 


44.9 




85.7 


sew 


OP 


9.8 


+ 


2.7 


7.50 


28.0 


4.6 




18.1 




CP 


14.1 


± 


4.4 


19.56 


31.3 


7.0 




24.5 




SP 


13.8 


+ 


3.1 


9.66 


22.5 


9.5 




17.6 


1,000-SW OP 


238.0 


± 


47.4 


2,243.00 


19.9 


170.0 




352.0 




CP 


188.0 


± 


31.3 


978.00 


16.6 


11.7 




26.5 




SP 


179.0 


± 


25.6 


656.94 


14.3 


14.7 




22.3 



'OP = open pollination; CP = controlled cross-pollination in a 10 x 10 diallel; SP = self 
pollination of 10 parents. 



Similar data can be foimd for cone diameter, seeds per 
cone, seeds per cone weight, and 1,000-seed weight in 
table 4. 

CORRELATIONS 

Significant (p < 0.05) and highly significant (p < 0.01) 
correlations were found between cone length and cone di- 
ameter, cone length and seeds per cone, cone length and 
seeds per cone weight; cone diameter and seeds per cone, 
cone diameter and seeds per cone weight; seeds per cone 
and seeds per cone weight. No significant correlation 
was found between 1,000-seed weight and any other trait 
(table 5). These strong correlations among the main cone 
and seed traits suggest that improvement (quantitatively) 
of seed production could be made by indirect selection; for 
example, selection for cone length, as an easily measur- 
able trait, will cause an increase in seeds per cone and 
seeds per cone weight, and consequently in seed production. 

TYPE OF DISTRIBUTION 

The distribution frequencies of cone length, cone diam- 
eter, seeds per cone, seeds per cone weight, and 1,000- 
seed weight were very close to normal distribution (fig. 2). 
According to genetic theory (Mather and Jinks 1977), this 
pattern of distribution is specific to quantitative traits. 
Such traits are polygenically controlled. 



156 



Table 5 — Phenotypic correlations among Pinus cembra cone and 
seed traits (Df = 1 8) 



Trait 


CL 


CD 


SC 


sew 


1 nnn.sw 


CL 


1.000 


0.577- 


0.636- 


0.669- 


0.272 


CD 




1.000 


0.499* 


0.647- 


0.227 


SC 






1.000 


0.636- 


0.325 


sew 








1.000 


0.396 


1,000-SW 










1.000 



DISCUSSION 

Our observations in the main Carpathian populations 
indicated that tree, locahty, and year in which the cones 
were initiated may significantly affect cone and seed 
traits. As the measurements were performed on cones 
collected in two different years, the comparisons were 
perhaps not entirely valid. 



The seeds per cone and seeds per cone weight from open- 
pollinated cones ranked lower than the same parameters 
from both cross-pollinated and self-pollinated cones; but, 
surprisingly, 1,000-seed weight from open-pollinated 
cones ranked higher than 1,000-seed weight from cross- 
pollinated cones. The lack of significant correlation 
between 1,000-seed weight and seeds per cone weight 
(table 5) partially explains this unexpected resvilt. Also, 
these differences could be attributable to the biological 
and climatic factors (temperature, moistm-e, and wind) 
that occurred in 1990 and 1991. 

The cone length and cone diameter means from self- 
pollinated cones were similar to cone length and cone 
diameter means from both open-pollinated and cross- 
pollinated cones, but the seeds per cone and seeds per 
cone weight means from self-pollinated cones were similar 
to seeds per cone and seeds per cone weight means from 
cross-pollinated cones and ranked higher them open- 
pollinated cones. Consequently, stone pine was found 



£ 20 




30 



^ 20 - 



O 



£ 10 




330 



3.7 3.9 4.2 4.5 4.8 5.1 5.4 5.7 6.0 

Cone Length (cm) 



3.1 3.4 3.7 4.0 4.3 4.6 4.9 

Cone Diameter (cm) 



^ 20 



0> 



S> 10 




10 20 30 40 50 60 70 80 90 

Seeds per Cone (No.) 



30 - 



^ 20 ■ 



o 



2> 10 




4 7 10 13 16 19 22 



>g 20- 



0) 



0) 




150 190 230 270 310 350 390 430 



Seeds per Cone Weight (g) 1 ,000-Seed Weight (g) 

Figure 2— Frequency distribution in the Pinus cembra populations evaluated for five traits. 



157 



to be highly self-fertile. This led to the conclusion that 
a very high proportion of the seeds produced by wind pol- 
lination would be selfs; this is an undesirable characteris- 
tic since selfed seed produces slower growing trees and 
the seedlings have a lower survival rate. 

Finally, it should be stressed that although the stone 
pine cone diameter, cone length, and seeds per cone 
means from the Alps exceeded the same trait means from 
the Carpathians, the 1,000-cone weight mean from the 
Carpathians was similar to that from the Alps. 

CONCLUSIONS 

Southern populations were similar in cone length and 
cone diameter but not in seeds per cone, seeds per cone 
weight, and 1,000-seed weight. Southern populations 
ranked high and the northern ones ranked low in all 
measured traits. 

Very high variation in both seeds per cone and seeds 
per cone weight was found within each population, but 
variation in 1,000-seed weight was moderate, and it was 
low in cone length and cone diameter traits. 

Cone length and cone diameter from open-pollinated, 
cross-pollinated, and self-pollinated cones were similar, 
while seeds per cone, seeds per cone weight, and 1,000- 
seed weight were not. 

The strong correlations among the main cone and seed 
traits suggest that genetic improvement in seed produc- 
tion could be attained by indirect selection. 

All measured traits displayed a continuous variation, 
suggesting polygenic control. 

ACKNOWLEDGMENTS 

Gratitude is expressed to Dr. H. Kriebel from Ohio 
State University for the revision of the English version 
of the paper and for the useful suggestions. 

REFERENCES 

Beldie, A. 1941. Observatii asupra vegetatiei lemnoase 
din Muntii Bucegi. Analele ICEF, Seria 1, vol. 6: 39-43. 

Blada, I. 1982. Relative blister rust resistance of native 
and introduced white pines in Romania. In: Heybroek, H.; 
[and others], eds. Resistance to diseases and pests in 
forest trees. PUDOC, Wageningen, The Netherlands: 
415-416. 

Blada, I. 1987. Genetic resistance to Cronartium ribicola 
and height growth in some five needle pines and of 



some inter-specific hybrids. Bucharest: Academy of 
Agricultural and Forestry Sciences. 146 p. Thesis. 

Blada, I. 1990a. Breeding Pinus cemhra by intra- and 
inter-specific crosses. For. Res. Inst., Bucharest, Re- 
search Plan, No. 7/1990. 

Blada, I. 1990b. Genetic variability of some traits in two 
Pinus cembra natural populations. In: Garrett, P. W., 
ed. Proceedings of a s3miposium on white pines prov- 
enances and breeding. Gen. Tech. Rep. NE-155. Radnor, 
PA: U.S. Department of Agriculture, Forest Service, 
Northeastern Forest Experiment Station: 56-68. 

Blada, I. 1992a. Pinus cembra and its inter-specific hy- 
brids. Paper presented at the lUFRO Centennial Meet- 
ing, Berlin-Eberswalde, August 31-September 5, 1992. 

Blada, I. 1992b. Diallel cross in Pinus cembra. I. Variation 
in cone and seed traits. [In preparation]. 

Contini, L.; Lavarelo, Y. 1982. Le pin cembro: repartition, 
ecologie, sylviculture et production. Paris: INRA, ISBN. 
197 p. 

Critchfield, W. B.; Little, E. L., Jr. 1966. Geographic dis- 
tribution of the pines of the world. IVIisc. Publ. 991. 
Washington, DC: U.S. Department of Agriculture, For- 
est Service. 97 p. 

Griffing, B. 1956. Concept of general and specific com- 
bining ability in relation to diallel crossing systems. 
Australian J. Sci. 9: 465-493. 

Gubesch, L. 1971. Raspandirea relictului glaciar zambrul 
(Pinus cembra) pe versantii sudici ai unor masive din 
Calimani. Ocrotirea Naturii, 15, 2: 149-159. 

Holzer, K. 1972. Intrinsic qualities and growth potential 
of Pinus cembra and P. peuce in Europe. In: Bingham, 
R. T.; [and others], eds. Biology of rust resistance in for- 
est trees. Misc. Publ. Washington, DC: U.S. Department 
of Agriculture, Forest Service: 99-110. 

Holzer, K. 1975. Genetics of Pinus cembra L. Anales 
Forestales, Zagreb. 6/5: 139-158. 

Mather, K.; Jinks, J. L. 1977. Introduction to biometrical 
genetics. Chapman and Hall, Ltd. 233 p. 

Moser, L. 1960. Verbreitung und Bedentung der Zirbe 
im italienishcen Alpengebiet. Jb. Ver. Z. Schutz 
Alpenpfl.u.-tiere. 25: 16-21. 

Oarcea, Z. 1966. Contributii la cunoasterea raspandirii si 
vegetarii pinului cembra in Retezat. Revista Pad. 9: 
495-497. 

Rohmeder, E.; Rohmeder, M. 1955. Untersuchungen uber 
das Samentragen und Keimen der Zirbelkiefer (Pinus 
cembra) in den Bayrischen Alpen. Allg. Forstzeitschrift. 
10: 83. 

Tataranu, I. D.; Costea C. 1952. Un arbore de interes 
forestier: Pinus cembra. Revista Padurilor. 11: 3-14. 



158 



ROLE OF NUTCRACKERS ON SEED 
DISPERSAL AND ESTABLISHMENT OF 
PINUS PUMILA AND P. PENTAPHYLLA 

Mitsuhiro Hayashida 



AbBtract— Nutcrackers (Nucifraga caryocatactes) harvest almost 
all Pinus pumila Kegel cones and cache the seeds in the soil. Most 
P. pentaphylla 'M.ayr seeds scatter after the cones open, while nut- 
crackers collect the seeds still held in the cones and then cache the 
seeds. Many seedlings of P. pumila and P. pentaphylla are often 
found in clusters consisting of several individuals of the same age. 
Almost all P. pumila seedlings and clusters of P. pentaphylla 
seedlings on barren slopes likely originate from seeds cached by 
nutcrackers. 



Stone pines (subsection Cembrae in Pinus) have large 
wingless seeds that are dispersed by caching of vertebrates 
(Hayashida 1989a; Hutchins and Lanner 1982; Mattes 
1982; Tomback 1982). Nutcrackers {Nucifraga spp.) are 
the main dispersal agents of seeds of these pines (Lanner 
1989). The nutcracker also caches other large wingless 
seeds or short-winged seeds in section Strobus (Lanner 
and Vander Wall 1980; Tomback 1990; Vander Wall and 
Balda 1977). 

Pinus pumila Regel (Japanese stone pine) is a dwarf pine. 
Its range extends through Japan and Korea into Siberia, 
Kamchatka (Critchfield and Little 1966). On most of the 
high mountains in Japan, a vegetational zone dominated 
by P. pumila occurs above the forest limit. This area is 
called the "Pinus pumila zone" by Japanese ecologists. The 
regeneration of P. pumila thickets is mainly due to a veg- 
etative regeneration by adventitious roots (Okitsu and Ito 
1983). Large wingless seeds of P. pumila are known to be 
dispersed by nutcrackers (Saito 1983; Turcek and Kelso 
1964), but a detailed observation for seed dispersal and 
caching behavior is needed. 

Pinus parviflora Sieb. et Zucc. (Japanese white pine) is 
distributed throughout the Japanese islands (Critchfield 
and Little 1966). This species consists of two geographical 
varieties, which intergrade in central Japan (Hayashi 1954). 
The northern variety is sometimes called P. pentaphylla 
Mayr; the southern P. himekomatsu Miyabe and Kudo or 
a variety of the former. Since the seed wings of P. penta- 
phylla are longer than those of P. himekomatsu (Ishii 1968), 
I distinguished between the two pines. There are few eco- 
logical studies of these pines except for community struc- 
ture (Tatewaki and others 1960; Yoshioka and Saito 1962). 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Mitsuhiro Hayashida is Associate Professor, Department of Bioenviron- 
ment, Faculty of Agriculture, Yamagata University, Wakaba, Tsuruoka 
997, Japan. 



Pinus pumila and P. pentaphylla are in the subgenus 
Strobus and are so similar that they naturally hybridize 
(P. hakkodensis Makino) (Ishii 1968); but these pines differ 
in seed, cone, and other characteristics. The objectives of 
this paper are to describe the processes from seed dispersal 
to seedling estabUshment of P. pumila and P. pentaphylla 
and to examine the role of nutcrackers in their regeneration. 

STUDY AREA 

The study was conducted on Mount Apoi (811 m above sea 
level), which is located in south-central Hokkaido, Japan 
(42°6' N., 143°2' E.). It is composed of ultramafic rocks 
(dunite, IherzoHte, etc.) (Niida 1984), and is covered with 
coniferous forests. The P. pumila zone extends from eleva- 
tions of 500 m to 800 m. Below 500 m, P. pentaphylla for- 
est occupies the rocky slopes and ridges, while the greater 
part of the mountain is covered by coniferous forests domi- 
nated by Picea glehnii, Pinus pentaphylla, and Abies 
sachalinensis (Hayashida 1989b). 

METHODS 

The number and distribution of 2-year cones were mapped 
to determine rates of seed predation by vertebrates and in- 
sects. These cones were counted from July to October in 
1984-87. The observations on seed harvesting, transport- 
ing, caching, and retrieving behavior in vertebrates were 
made in 1984-87 using binoculars and a telescope. 

Many seedlings of P. pumila and P. pentaphylla were 
found on barren slopes along mountain trails. These seed- 
lings were often found in clusters, which consisted of sev- 
eral seedhngs of the same age. To obtain the frequency 
distribution of number of individuals per cluster in both 
pines, all seedlings on barren slopes along mountain trails 
were recorded. 

SEED AND CONE MORPHOLOGY 

Pinus pumila has wingless seeds; P. pentaphylla seeds 
have wings that are seed length (table 1). Seeds of both 
pines have the same thick coats. Pinus pumila seeds are 
significantly heavier than those of P. pentaphylla, though 
the seeds of P. pentaphylla are larger. 

Pinus pumila cones do not open when they ripen, but 
P. pentaphylla cones open at maturity and release seeds 
to fall free. Pinus pumila cones weigh less than P. penta- 
phylla cones, but contain more seeds. 



159 



Table 1 — Seed and cone characteristics of Pinuspumila and P. pentaphylla 



Pinus pumila P. pentaphylla 



MAAQIirOITIAnt 

d 1 Id It 


Moan + Qn 


M 


Moon 4- QD 


Kl 

n 


P< 


Seed length (mm) 


7.700 ± 0.600 


105 


10.000 ± 0.700 


105 


0.001 


Dry weight (g) 


0.100 ± 0.014 


36 


0.068 ± 0.014 


20 


0.001 


Seed coat thickness (mm) 


0.390 ± 0.040 


20 


0.390 ± 0.040 


20 


NS 


Length of wings (mm) 


wingless 




10.500 + 1.600 


105 




Cone mass (g) 


11.00013.300 


32 


24.800 ± 7.000 


27 


0.001 


Number of seeds per cone 


42.800 ±12.900 


45 


28.900 ±9.100 


38 


0.001 


Dehiscence 


not open 




open 







'Mann-Whitney U test. 



SEED DISPERSAL OF P. PUMILA 

Figure 1 shows the seasonal disappearance of 2-year 
cones of P. pumila in 1985-88. All cones disappeared from 
trees by mid-October every year. I observed nutcrackers 
(Nucifraga caryocatactes), squirrels (Sciurus vulgaris ori- 
entis), and chipmunks (Tamias sibiricus lineatus) harvest- 
ing cones. Five species of vertebrates harvested and car- 
ried seeds to caches: nutcracker, vsiried tit (Parus varius), 
nuthatch (Sitta europaea), squirrel, and chipmunk. The 
nutcracker was the most frequently observed vertebrate 
carr3dng P. pumila seeds. 

Nutcrackers usually harvest seeds by removing the 
cones and flying with them to perches where they extract 
the seeds and hold them in their sublingual pouches. Var- 
ied tits and nuthatches were able to take P. pumila seeds 
exposed by nutcracker foraging and eat them. They were 
seen caching pine seeds on a tree or in the soil. Squirrels 
visited the P. pumila scrub from a coniferous forest to for- 
age on pine seeds. They often carried a whole cone to cache 
it in scatter-hoarding. Chipmunks harvest pine seeds in 
their cheek pouches. They carried about 50 seeds on a trip 
and cached them in the soil. 

These results indicate that nutcracker is the main seed 
disperser of P. pumila. 




(0 

E 
« 
oc 





Jul. 


Aug. 


Sep. 


Oct. 



Figure 1 — Retention of mature cones of Pinus pumila 
on trees (in years 1985-88) as related to season. 



SEED DISPERSAL OF 
P. PENTAPHYLLA 

Cones taken by vertebrates or damsiged by insects 
(cone-boring lepidopteran larvae) accounted for less than 
20 percent of the total. Usually, nearly 90 percent of the 
cones open on the trees from mid-September through early 
October. After opening, about 80 percent of the seeds im- 
mediately scatter to the ground, due to the fact that 77 
percent of P. pentaphylla cones were oriented downward 
(Hayashida 1989b). 

Nutcrackers, varied tits, and nuthatches harvested and 
cached seeds still held in the upward-oriented open cones. 
In the years when P. pumila crop sizes were small, nut- 
crackers harvested P. pentaphylla seeds more frequently. 
They extracted the several seeds from an open cone, peeled 
off the seed wings, and put seeds into their sublingual 
pouches. They moved rapidly around a tree and went from 
tree to tree. 

NUTCRACKER CACHING BEHAVIOR 

Nutcrackers began to cache P. pumila seeds in early 
August when the seeds were mature, and they continued 
to harvest seeds until the seed crop was depleted. They 
cached the seeds in the soil at a depth of 2 to 3 cm. Cache 
size (the number of seeds per cache) ranged from 1 to 51. 
Nutcrackers cached P. pumila seeds mainly in coniferous 
forests. 

In mid-September, P. pentaphylla cones began to open, 
then nutcrackers began to harvest seeds from the opening 
cones. Nutcrackers scatter-hoarded almost all seeds in the 
soil (97 percent) except for two caches containing one and 
two seeds that were established in the bark on fallen trees. 
Cache size ranged from 1 to 40 seeds. Seventy-eight caches 
were recorded in coniferous forests and only one cache in 
the Pinus pumila zone. 

CLUMPS OF SEEDLINGS 

Eighty-four percent of the occurrence of P. pumila seed- 
lings were in groups of two or more trees (table 2). These 
seedlings were 1 km in distance from the nearest seed tree. 



160 



Table 2 — Number of individuals per clump of Pinus pumila and P. pentaphylla on 
barren slopes along mountain trails 





Number of clumoQ 


Numbor of inHiviclijalQ 




(percent) 


per clump 


Species 


Single Clump (2<) 


Maximum Mean ± SD 


Pinus pumila 


9(16) 47(84) 


26 5.6 ±5.3 


P. pentaphylla 


137(72) 54(28) 


25 2.6 ± 3.8 



Twenty-eight percent of P. pentaphylla seedlings were 
growing in groups of two or more trees. There were many 
seed trees of P. pentaphylla along the moimtain trails. Sin- 
gle seedlings might originate from naturally scattered fallen 
seeds, and the seedlings in cliunps probably originated from 
caches by nutcrackers. 

DISCUSSION 

All mature cones of P. pumila are harvested by verte- 
brates every year (fig. 1). My observations indicate that 
almost all of the transported seeds were carried by nut- 
crackers. Nutcrackers cached seeds in the soil at a depth 
of 2 to 3 cm. It is shallow enough to permit seedling estab- 
lishment. Therefore, seed dispersal of P. pumila is heavily 
dependent on seed caching by nutcrackers. 

Nutcrackers carried and cached P. pumila seeds mainly 
into coniferous forests. In general, P. pumila seedhngs can- 
not grow and establish in coniferous forests even if cached 
seeds survive and germinate. However, several P. pumila 
trees were found that developed and matured on the rocky 
ridges or barren slopes in coniferous forests. These facts 
may indicate that the nutcracker is not an efficient seed 
disperser of P. pumila under the present conditions. Pinus 
pumila adapted its specialized tree form and other features 
to the habitat that is characterized by strong winds and 
heavy snow accumulation in winter. Thus, the P. pumila 
zone is developed on deforested areas in the boreal subal- 
pine belt. In Europe, nutcrackers cached P. cembra seeds 
in coniferous stands and sometimes cached them above 
timberline (Mattes 1982). Seed dispersal over wide areas 
by nutcrackers probably played an important role in ex- 
panding the range of P. pumila. 

Pinus pentaphylla has dehiscent cones and winged seeds. 
Most of the seeds scattered when cones opened. Nutcrack- 
ers often harvested and cached seeds still held in the up- 
ward open cones. About 90 percent of P. pentaphylla trees 
found in coniferous forests were individual single trees, but 
on the ecotone between coniferous forest and the Pinus 
pumila zone, half of the occurrences of trees were in clumps 
of two or more individuals (Hayashida 1989b). About 30 
percent of the seedlings were growing in clumps on barren 
slopes in the coniferous forest zone. These seedlings in 
climips probably originated from seeds cached by nutcrack- 
ers. Seed dispersal by nutcrackers enables P. pentaphylla 
to invade areas that are inaccessible to other coniferous 
trees. 

The northern variety, P. pentaphylla, has longer seed 
wings than the southern variety, P. himekomatsu (Ishii 
1968). It is possible that these two varieties have different 



morphological features and seed-dispersal syndromes. If 
so, why? I would study these problems with an ecological 
approach. 

ACKNOWLEDGMENTS 

I thank Urakawa Local Forest Management Office, 
Hokkaido, and Oji Forestry and Landscaping Co. for per- 
mission to work in the area and for information on the 
area. 

REFERENCES 

Critchfield, W. B.; Little, E. L., Jr. 1966. Geographic distri- 
bution of pine of the world. Misc. Publ. 991. Washington, 
DC: U.S. Department of Agriculture, Forest Service. 97 p. 

Hayashi, Y. 1954. The natural distribution of important 
trees, indigenous to Japan. Conifers Rep. 3. Bulletin of 
the Government Forestry Experiment Station. 75: 1-173. 
[In Japanese with English svmimary]. 

Hayashida, M. 1989a. Seed dispersal by red squirrels and 
subsequent establishment of Korean pine. Forest Ecology 
and Management. 28: 115-129. 

Hayashida, M. 1989b. Seed dispersal and regeneration 
pattern of Pinus parviflora var. pentaphylla on Mt. Apoi 
in Hokkaido. Research Bulletins of the College Experi- 
ment Forests, Facvdty of Agriculture, Hokkaido Univer- 
sity. 46(1): 177-190. [In Japanese with English stmimary]. 

Hutchins, H. E.; Lanner, R. M. 1982. The central role of 
Clark's nutcracker in the dispersal and establishment 
of whitebark pine. Oecologia. 55: 192-201. 

Ishii, S. 1968. The basic forestry study on genus Pinus, 
especially its taxonomic and geographical consideration. 
Journal of the Faculty of Agriculture, Kochi University. 
19: 1-114. [In Japanese with English svunmary]. 

Lanner, R. M. 1989. Biology, taxonomy, evolution, and 
geography of stone pines of the world. In: Schmidt, 
Wyman C; McDonald, Kathy J., comps. Proceedings — 
symposiimi on whitebark pine ecosystems: ecology and 
management of a high-mouintain resource; 1989 March 
29-31; Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, 
UT: U.S. Department of Agriculture, Forest Service, In- 
termountain Research Station: 14-24. 

Lanner, R. M.; Vander Wall, S. B. 1980. Dispersal of lim- 
ber pine seed by Clark's nutcracker. Journal of Forestry. 
78: 637-639. 

Mattes, H. 1982. Die Lebensbemeinschaft von Tannenhaher, 
Nucifraga caryocatactes (L.), und Arve, Pinus cembra L., 
und ihre forstliche Bedeutiing in der oberen Gebirgs- 
waltstufe, Berichte. 241. 74 p. 



161 



Niida, K. 1984. Petrology of the Horoman utramafic rocks 
in the Hidaka Metamorphic Belt, Hokkaido, Japan. Jour- 
nal of the Faculty of Science, Hokkaido University, Ser. 
IV. 21: 197-250. 

Okitsu, S.; Ito, K. 1984. Vegetation d3niamics of the Siberian 
dwarf pine (Pinus pumila Kegel) in the Taisetsu moun- 
tain range Hokkaido, Japan. Vegetatio. 58: 105-113. 

Saito, S. 1983. Caching of Japanese stone pine seeds by 
nutcrackers at the Shiretoko Peninsula, Hokkaido. Tori. 
32: 13-20. [In Japanese with English summary]. 

Tatewaki, M.; Tsujii, T.; Kawano, S. 1960. The community 
structure and distribution of Pinus parvifLora forest in 
Hokkaido. Japanese Journal of Ecology. 10: 120-123. [In 
Japanese with English summary]. 



Tomback, D. F. 1982. Dispersal of whitebark pine seeds by 
Clark's nutcracker: a mutualism hypothesis. Journal of 
Animal Ecology. 51: 451-467. 

Tomback, D. F. 1990. The evolution of bird-dispersed 
pines. Evolutionary Ecology. 4: 185-219. 

Turcek, F. J.; Kelso, L. 1968. Ecological aspects of food 
transportation and storage in the Corvidae. Communi- 
cations in Behavioral Biography, Al: 277-297. 

Vander Wall, S. B.; Balda, R. P. 1977. Coadaptations of the 
Clark's nutcracker and the pinon pine for efficient seed 
harvest £md dispersal. Ecological Monographs. 47: 89-111. 

Yoshioka, K; Saito, K. 1962. Differences in the distribution 
of forest communities due to topography of the habitats 
in Okunikkawa, Miyagi Prefecture. Ecological Review. 
15: 213-220. 



162 



ROLE OF VARIOUS ANIMALS IN 
DISPERSAL AND ESTABLISHMENT 
OF WHITEBARK PINE IN THE ROCKY 
MOUNTAINS, U.S.A. 

Harry E. Hutchins 



Abstract — The dispersal and establishment of whitebark pine 
has been attributed to Clark's nutcracker. But what is the role 
of other animals who forage on whitebark pine seed? This study 
looks at the role of the diurnal birds and mammals that forage 
on whitebark pine cones and seed in western Wyoming. A total 
of 1,005 cones were carefully observed in both contiguous "forest" 
stands and open-grown "meadow" trees in early November. Red 
squirrels harvested most of the whitebark pine seed in forested 
plots while Clark's nutcracker removed most of the seed from the 
meadow trees. Only a small percentage of seed went to ground 
squirrels, chipmunks, Stellar's jays, ravens, etc. The foraging 
behavior as well as the importance of these animal species to 
whitebark pine regeneration is discussed. 



Many animal species have been observed foraging on 
the nutritious seeds of whitebark pine (Pinus albicaulis). 
But exactly which of these animals plays a role in seed- 
ling establishment and to what degree? Which animals 
are seed predators? Can whitebark pine seed germinate 
from that seed which is dropped through animal foraging 
accidents? Do animals harvest all the seed in the fall or 
can cones fall to the ground and seed germinate as the 
cone decays? This paper will address these questions. 

The role of various animals in the dispersal and estab- 
lishment of whitebark pine was examined by observing 
1,005 cones and determining where and how the seeds 
were dispersed. By primarily observing the behavior of 
diurnal birds and mammals in the subalpine canopy, I 
was able to determine the degree of influence each species 
has on seed dispersal and seedling establishment. 

METHODS 

The study was primarily conducted at the Squaw Basin- 
Togwotee Pass area (Bridger-Teton National Forest; fig. 1). 
This area of high-elevation meadows offered distant views 
of bird activity, whitebark pine growing in both contigu- 
ous forest and open-grown situations, and an abundant 
cone crop for whitebark pine diuing 1980. Besides white- 
bark pine, the forest stands were composed of Engelmann 
spruce (Picea engelmannii), subalpine fir {Abies lasio- 
carpa), and a minor amoimt of lodgepole pine (Pinus con- 
torta). The moraine meadow ridges were pioneered by 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Harry E. Hutchins is Natural Resource Instructor, Itasca Community 
College, Grand Rapids, MN. 



lone whitebark pine. This provided an opportimity to 
study two different types of commimities — "forest" and 
"meadow" (fig. 2). 

Observations of animal activity were also made at 
Mount Washburn (2,680 to 3,140 m, Yellowstone National 
Park, WY) and Surprise Lake (2,960 m. Grand Teton Na- 
tional Park, WY). These areas were similar to the Squaw 
Basin site except they lacked open-grown, cone-bearing 
trees. Details on the study sites can be foxmd in Hutchins 
and Lanner (1982) and Vander Wall and Hutchins (1983). 

Several cone-bearing whitebark pines were chosen for 
observation in both continuous forest in which red squir- 
rels {Tamiasciurus hudsonicus) were present and in the 
open meadow, at least several hundred meters from the 
nearest forest edge (these sites lacked squirrels). In each 
of these two types of sites, 1,005 whitebark pine cones 
were observed during the period Jvdy 3 to November 2, 
1980. The cones were scattered among several mature 
trees and were counted at 1- to 2-week intervals using 
a 15-25 X 60-mm spotting scope. Coimts were made by 
standing in a specific marked location and "mapping" the 
cones on clear acetate. Changes in the "cone map" at each 
observation were recorded, including the partial removal 
of a cone. Cone count data were converted into seeds us- 
ing an empirically derived value for the mean number of 
seeds per cone (50.4 seeds/cone, Hutchins 1982). Partially 
consimied cones were tallied by estimating from the ground 
the percentage of seed remaining as described in Hutchins 
and Lanner (1982). The seed harvest data were then plot- 
ted against ciunulative time for both forest and meadow 
sites. At each observation date samples of seeds were col- 
lected to determine maturity and condition. The numbers 
of filled, discolored, insect-attacked, and second-year- 
aborted ovules were tallied. Mean dry weight of shelled 
seeds, seed coat thickness, and caloric content of shelled 
seeds were all obtained. Germination tests were also con- 
ducted with seeds collected at 10 collection dates during 
the 1980 field season. 

Seedfall (caused by animal foraging) below the tree was 
estimated by placing five, 1-m^ vnre mesh seed traps ran- 
domly below trees. The tops of the seed traps were de- 
signed to let seed fall through but to exclude rodents. 
Data on the nimiber of cones in the tree, the area of the 
tree crown, and the number of seeds falling per square 
meter were used to estimate the magnitude of seed fall. 

Predation on seed caches of whitebark pine was also 
studied in 1979 and 1980. I simulated three types of 
caches: (1) seed that lands on the soil surface from 
foraging accidents; (2) seed cached at a depth of 3 cm 



163 




MONTANA 



WYOMING 



GRAND TEfoN 
NAT.' PARk 

I i ® 

Surprise Lake ^ 

i A 



SHOSHONE 
NATIONAL 
FOREST 



LEGEND 



— PAVED ROAD 
® STUDY SITE 

KILOMETERS 



WYOMING 



simulating a Clark's nutcracker (Nucifraga columbiana) 
cache; and (3) seed cached 7 cm deep, which simulates 
the most shallow red squirrel seed cache. Each individual 
cache contained 10 seeds. 

Animal time budget data were collected for the diurnal 
species found foraging in the whitebark pine ecosystem. 
Both quantitative and qualitative observations were made 
on the various activities and behaviors of the animals; the 
methods are detailed in Hutchins and Lanner (1982). 

SEED DEVELOPMENT 

Hutchins and Lanner (1982) monitored seed develop- 
ment and found mean seed coat thickness and seed weight 
to be significantly greater during the August 31 to 
November 2 collecting period than prior to these dates. 
Clark's nutcrackers were unable to extract whole seed 
from a cone ;intil August 13 due to the thin, fragile seed 
coats. The characteristic shell fragments of an unripened 
"nutcracker cone" were left as evidence of their foraging 
attempts. Thus, nutcrackers were unable to cache whole 
seed until after this date. 

By September 7-10, the cones had dried and turned a 
dull brown from their previous moist, pitch-filled, purple 
color. Whitebark pine cones are often referred to as inde- 
hiscent; however, about 25 percent of the cones collected 
after September 7 parted their scales slightly (4-8 mm; 
n = 141). This still was not enough of an opening to allow 
the seeds to fall out on their own accord. 



Figure 1 — Locations of three study sites in Wyoming. 




164 



SEED CROP DEPLETION 

Figure 3 (from Hutchins and Lanner 1982) shows seed 
crop depletion followed a logistic curve (r^ = 0.99 in the 
forest and = 0.96 in the meadow). Seed harvesting be- 
gan somewhat earlier in the forest than in the meadow. 
In fact, about 50 percent of the seed crop had been har- 
vested in the forest by August 31, while in the meadow 
this point was not attained until September 25. As seen 
on the graph, no seeds remained from this mast year in 
the forest stands by November 1. Only 0.1 percent of the 
seeds still resided in the tree crowns of the meadow trees, 
and these were no longer there when checked on 27 June, 
1981. 

Unless aided by an animal, cones of whitebark pine do 
not fall off a tree to the ground except rarely during ex- 
tremely large cone crops — as occurred in 1989 (Hutchins 
and Lanner 1982; Lanner 1982; Mattson and Reinhart, 
these proceedings). The following incident shows the ex- 
tent Clark's nutcracker will go to acquire seed. Hutchins 
and Lanner (1982) placed double-layer wire hardware 
screening over several cones so there would be cones to 
collect for germination studies. By late October — after 
available whitebark pine seeds had become scarce — the 
nutcrackers had ripped open the screening and harvested 
the seed. Seed that does drop to the ground from foraging 
accidents made up only 4.2 percent of the seed crop and 
most of this seed (69 percent) was determined unviable 
through examination of the contents in the shell during 
1980 (Hutchins 1990). 

ANIMAL INTERACTIONS WITH SEED 

A variety of diurnal animals were observed active in or 
imder whitebark pine trees. Many of these, however, were 
never observed foraging on whitebark pine seed in this 
study. These include: gray jay {Perisoreus canadensis). 



common flicker (Colaptus auratus), Cassin's finch (Cora- 
podacus cassinii), rosy finch (Leucosticte sp.), pine siskin 
(Spinas pinus), dark-eyed junco (Junco hymelis), black 
billed magpie (Pica pica), pine marten (Martes americana), 
coyote (Canis latrans), and weasel (Mustela sp.). 

Many feces of the mammal species were examined, but 
no evidence of whitebark pine seeds was found. Gray jays 
were commonly found in the crowns of whitebark pine but 
were only observed hawking insects and caching fresh 
carrion or boli in pine branches. Dow (1965) found this 
species has little interest in pine seed during feeding tri- 
als, although Turcek and Kelso (1968) describe its Eur- 
asian cousin, the European jay (P. infaustus), as having 
been observed taking and caching Siberian stone pine 
(Pinus sibirica). 

Of the birds, only the magpie could potentially be in- 
volved in the dispersal of whitebark pine (Smith and Balda 
1979). It is, however, very rare in the subalpine zone dur- 
ing the maturation of whitebark pine seed (Hutchins and 
Lanner 1982); consequently, any potential harvesting of 
seed would be unimportant to the regeneration of white- 
bark pine. 

Many species have been observed foraging on whitebark 
pine cones (Hutchins 1990, table 2). At this point I would 
like to take a closer look at their role in the dispersal and 
establishment of this pine. 

BIRD DISPERSERS 

Clark's Nutcracker — Nutcrackers were the most 
common resident bird to visit the whitebark pine trees 
(Hutchins and Lanner 1982). They were scattered about 
whitebark pine stands in loose flocks foraging and caching 
seed of this tree. They are dependent on these caches 
year round (Giuntoli and Mewaldt 1978; Vander Wall 
and Balda 1977; Vander Wall and Hutchins 1983). 




August 



September 



October 



November 



Figure 3 — Seasonal course of 
whitebark pine seed harvest by 
vertebrates in Squaw Basin, 
WY, in 1980. 



165 



Nutcrackers were observed harvesting seed as early 
as July 13, 1979. During July, however, they appeared to 
only be testing the cones for ripeness and primarily feed- 
ing on the previous year's caches until mid-August. Seed 
harvested at this time were lost to regeneration because 
the seed coats were broken and the seed is ungerminable 
vintil August 13. In 1980, the birds were able to success- 
fully harvest whole, developed seed by August 15. Nut- 
crackers harvested seed from cones at increasingly faster 
rates (Hutchins and Lanner 1982) through early October, 
when seed became hard to find (fig. 3). Nutcrackers were 
never seen attempting to harvest Engelmann spruce or 
lodgepole pine, even as the whitebark pine seed crop dwin- 
dled in October. As the seed supply became scarce in mid- 
October, I observed one bird checking over 50 cones for a 
period of 613 seconds without finding a single seed. 

By November 2, it appears nutcrackers were almost 
totally dependent on their new seed caches, and would 
remain so imtil the following August. These caches were 
placed just below the soil surface (2-3 cm), and over 70 per- 
cent of the caches were either one- two- or three-seed 
caches (x = 3.2 + 2.8 seeds/cache; fig. 4). The largest cache 
was 14 seeds. These data agree closely with Tomback 
(1978) and Vander Wall and Balda (1977), although the 
number of single-seeded caches was higher. 

Clark's nutcrackers cached their seeds in almost every 
conceivable soil type and microsite in the various study 
sites. Specific examples were given in Hutchins (1990). 
Although they may cache seeds on a wide variety of sites 
in the Rocky Mountains, several studies have indicated 
that south-facing slopes seem to be preferred (Lanner 
1982; Lanner and Vander Wall 1980; Snethen 1980; 
Tomback 1978; Vander Wall and Balda 1977). Seedling 
establishment appears to be much more common on more 
moist sites in the Rocky Mountains (Arno 1986; Arno and 
Hoff 1989; Vander Wall and Hutchins 1983) than in the 
Sierra Nevada (Tomback 1982). 



40- 



0) 

o 

(0 

O 



c 

o 
o 

Q. 



35 
30 
25 
20 
15 
10 

5 ^ 



1 2 3 4 5 6 7 8 9 10 11 12 13 14 

Number of Seeds per Cache 

Figure 4 — Clark's nutcracker cache size fre- 
quency. Data from 1 57 observations made at 
several locations in Wyoming. 



Transport distances varied greatly. Seeds were placed 
as close as 50 m from the site of the harvested tree, or 
were transported at least 3.5 km to the Breccia Cliffs on 
the edge of the Squaw Basin site. With little seed left in 
the trees by mid-October, the birds began retrieving many 
of their caches made in the Squaw Basin meadows and 
recaching them on the southwest-facing slopes of the 
Breccia Cliffs. 

Large flocks of nutcrackers would often cache seed 
together. At Mount Washburn, a flock estimated at 
150 birds was seen caching seed under an open-grown 
whitebark pine stand, with 10-15 birds within a 10-m^ 
area. No aggression occurred among the birds during 
these observations. 

By November 2, 1980, snow covered many of the cach- 
ing areas for the winter. Nutcrackers were observed on 
several occasions successfully pecking through up to 
25 cm of snow and ice to retrieve a cache during the win- 
ter months. The Togwotee Pass area may receive up to 
1,500 cm of snow a year, yet the windswept ridges and 
south-facing slopes remain exposed enough for the birds 
to retrieve their caches. Caches on the northeast-facing 
slopes and under the forest canopy are more frequently 
used during June, July, and August, as the snow recedes 
from these sites last (Vander Wall and Hutchins 1983). 

The number of seed an individual nutcracker caches 
annually has been estimated in several studies. The 
numbers for whitebark pine range from 32,000 in the 
Sierras (Tomback 1982) to 98,000 in the Rocky Moun- 
tains (Hutchins and Lanner 1982). Because of the num- 
ber of variables that must be considered (such as flight 
distances and the amount of available seed), these esti- 
mates vary a great deal from site to site and year to year. 

Studies by Tomback (1982), Vander Wall (1988), and 
Vander Wall and Balda (1977), estimate that an indi- 
vidual bird caches several times more seed than it needs 
to survive through the winter and early spring. At that 
time other food items become available — although the 
nutcrackers continue to use their caches heavily until 
the new cone crop begins to mature (Vander Wall and 
Hutchins 1983). This leaves many unused caches for 
potential germination and establishment of whitebark 
pine seedlings (Hutchins and Lanner 1982). 

Clark's nutcracker stands out from all the other poten- 
tial seed dispersers in two major ways. First, it consis- 
tently disperses seed in a way that increases the chance 
of seedling establishment. The seed is placed just below 
the soil surface and hidden from seed predators (Hutchins 
and Lanner 1982; Lanner 1980). 

Second, nutcrackers scatter caches across the 
landscape — both long and short distances from the source 
trees. Dispersal distances of up to 22 km have been ob- 
served by Vander Wall and Balda (1977). Also, the scat- 
tering of thousands of their caches reduces predation. 

Steller's Jay (Cyanocitta stelleri) — Having many of 
the behavioral and physiological attributes of the Clark's 
nutcracker (Vander Wall and Balda 1981), the Steller's 
jay could have potential as a dispersal agent of whitebark 
pine. In this part of their range, they were found to be 
primarily solitary foragers and uncommon visitors to the 
whitebark pine forest. These birds did not forage for 
whitebark pine seed until early September when the 



166 



cones dried and the scales separated slightly. Their bill 
structure does not approach the efficiency of the nut- 
cracker in prying apart cone scales to get to the seed 
(Vander Wall and Balda 1981), thus they were often only 
able to extract seed from cones that had been exposed by 
nutcrackers. These jays harvested whitebark pine seed 
from the ground and canopy in less than half of their ob- 
served foraging time during the fall of 1980 and 1981 
(Hutchins and Lanner 1982). The seed was always placed 
between their feet and hammered with their bill to crack 
the hull; nutcrackers primarily cracked the seed between 
their mandibles. 

Occasionally Steller's jays pouched their seed in their 
elastic esophagus for later caching. They were observed 
caching whitebark pine seeds on seven occasions; how- 
ever, none of these were in the soil. The birds placed 
the pine seeds in the crotch of a tree, a densely foliated 
witches broom, and under dense lichen growth along a 
tree branch. The largest number of seeds I observed be- 
ing pouched at one time by this species was five, although 
data from Vander Wall and Balda (1981) indicate these 
birds could hold up to 32 seeds/pouchload. Because this 
species does not cache whitebark pine seed in the soil and 
has a seed foraging rate comparable to small passerines 
(Hutchins and Lanner 1982), it is a very improbable agent 
for seedling establishment. 

Raven (Corvus corax) — third corvid rarely observed 
foraging on whitebark pine seed was the common raven. 
These birds had a great deal of difficulty extracting seed 
from whitebark pine cones due to their thick, long bills. 
They dropped most of the seed when foraging (Hutchins 
and Lanner 1982). Caching of carrion was observed with 
this species, but no observations were made of seed cach- 
ing. In Asia, Reimers (1959 cited in Tiircek and Kelso 
1968) described observations of ravens caching Japanese 
{Pinus pumila) and Siberian stone pine (P. sibirica), al- 
though it was not clear in what type of substrate. 

Other Passerines — Pine grosbeaks (Pinicola enuclea- 
tor) primarily foraged on whitebark pine seed that had 
been exposed by Clark's nutcrackers breaking off cone 
scales. Their large conical beaks also enabled them to 
tear away cone scales to get at seed. Their foraging rates 
were very slow compared to nutcrackers due to their tech- 
nique and, as a consequence, they had little influence on 
the depletion of the whitebark pine cone crop. 

Grosbeaks are not known to cache food (Smith and 
Balda 1979; Vander Wall 1990). During my observations 
of this species, they cracked the seed coat and consumed 
the seed in the tree crown. It is highly unlikely that they 
covild pass an intact seed through their digestive tract. 

The moxmtain chickadee (Parus gambeli) and red- 
breasted nuthatch (Sitta canadensis) occasionally searched 
through the whitebark pine cones during September after 
the seeds were exposed by nutcrackers. Both these spe- 
cies almost always dropped the seed from the tree crown, 
as the seeds were too large for them to handle. Their un- 
successful foraging bouts contributed to the small percent- 
age of seed found on the soil surface, which was later con- 
sumed by other animals. Neither bird species was observed 
caching whitebark pine seed, although they are known to 



cache smaller seeded conifer species imder bark (Smith 
and Balda 1979). 

MAMMAL DISPERSERS 

Red Squirrel (Tamiasciurus hudsonicus) — Red 

squirrels are common denizens of whitebark pine forests. 
In fact they were the second most commonly observed ver- 
tebrate next to the Clark's nutcracker (Hutchins and Lan- 
ner 1982). These mammals actively defend their territo- 
ries (Smith 1968), and like nutcrackers, harvest whitebark 
pine cones and seeds as a fall and winter food source. Red 
squirrels did not occiir on the meadow sites. 

Red squirrels spent most of their foraging time (75.8 per- 
cent) on whitebark pine cones or recovering dropped 
whitebark pine seeds. This includes the time it takes 
to place the cones in the midden. Most of the rest of the 
observed foraging activities were collecting Engelmann 
spruce cones (11.3 percent) and harvesting seeds of herba- 
ceous plants (12.9 percent). 

Red squirrels began harvesting cones as early as nut- 
crackers harvested seeds (Jvily 13, 1979), but did so more 
intensely during July while the nutcrackers relied more 
heavily on the previous year's caches (fig. 3). Their forag- 
ing rate is much higher than that of the other mammals 
discussed in this paper because they usually harvest en- 
tire cones. Most of the time, they remove the cone by pull- 
ing it off the branch with their teeth, instead of cutting 
the subtending branch (Hutchins 1990). Thus, cone re- 
moval is likely to cause little change in tree growth form 
or cone production. 

Cone caching began on August 4 at Squaw Basin. 
Red squirrels often would leave cones on the groimd below 
the tree where they were cut for periods of up to 3 days 
before caching them in their midden. Caching of other 
conifer cone species began later with Engelmann spruce 
on August 18 in 1980; subalpine fir (September 11 in 
1980), and lodgepole pine (September 27 in 1980). All 
cones were stored in "middens" {n = 114). These middens 
can be quite extensive and are composed of many years of 
cone debris above the soil surface (Finley 1969; Reinhart 
and Mattson 1990; Smith 1970). Of the time spent cach- 
ing food, 61.4 percent was devoted to whitebark pine 
cones and 16.9 percent to whitebark pine seeds. Foraging 
on the cones of other conifers in the subalpine made up 
most of the rest of the caching time (7,304 seconds of ob- 
servation). Red squirrels also cached some mushrooms 
and herbaceous seeds. 

Squirrels were also found to cache whitebark pine seeds 
in their midden beginning aro\md September 16, 1980. 
These seed caches were placed 6.5- to 40-cm deep {n = 6); 
four of these observations were between 11 and 11.5 cm 
deep. The number of seeds per cache ranged from 14 to 
55 ix = 28.8; 19.2 seeds/cache; n = 4), although Kendall 
(1981) foimd up to 176 seeds in a single hole. I examined 
the seeds from two squirrel caches gind found all the seeds 
sound. 

Red squirrels actively chased Clark's nutcrackers from 
the trees above their territories and their middens. On 
two occasions, however, I watched nutcrackers steal whiter 
bark pine seeds from squirrel middens. 



167 



Chipmunk (Eutamias sp.) — Chipmtinks are uncom- 
mon visitors to whitebark pine tree crowns but do occur 
on both open-meadow and forested sites. Most of their 
time is spent on the ground, usually near cover plants like 
sagebrush (Artemisia tridentata). They preferred herba- 
ceous plant parts (lupine seed, grasses, etc.) on my subal- 
pine study sites until the third week of September when 
these plants died back and presimiably lost much of their 
nutritive value. At that time, they clumsily foraged on 
the remaining whitebark cone crop (-10 percent of the 
seed was left in the forest, -40 percent in the meadow). 
They more commonly foraged on the smsdl amount of seed 
below the trees, which had been dropped by any of the 
species discussed earlier. Tevis (1953) and Heller (1971) 
similarly foimd that in the Sierra Nevada chipmxmks 
devote little time to foraging on whitebark pine seed and 
cones. 

I found no evidence of chipmimks caching whitebark 
pine seed. Broadbooks (1958) found the larder cache 
depth of yellow pine chipmimks {Eutamias amoenus) to 
average 28 cm. They use these types of caches as a winter 
and spring food source when they are periodically aroused 
during hibernation (Vander Wall 1990). I examined chip- 
munk burrows to a depth of 20 cm without finding any 
evidence of them harvesting and caching whitebark pine 
seed. 

Yellow pine chipmunks are known to scatter-hoard 
smaller amounts of Jeffrey pine (Pinus jeffreyi) and pon- 
derosa pine {P. ponderosa) seed in shallower caches from 
1 to 25 mm in depth (Vander Wall 1992a). These species 
of chipmunks in these studies, along with other rodents, 
may retrieve many caches in the fall in the Sierra Nevada. 
They use the seed to stock their winter larders, but 
many seeds survive and germinate the following spring 
(Vander Wall 1992b). Although chipmunks may play 
a role in the afforestation of certain pines, there are no 
data that support this conjecture for whitebark pine. 
Vander Wall (1992a) provides strong evidence that Jeffrey 
pine may arise from chipmunks (Tamias amoenus, specio- 
sus, and quadramaculatus) scatter-hoarding seed. In 
Japan, Hayashida (1989) found Siberian chipmunks 
(T. sihiricus) were not an effective dispersal agent of 
Korean stone pine {P. koraiensis) in a plantation. In this 
study, the limited amount of seed these animals harvest 
(see seed crop depletion section of this paper) apparently 
precludes them from being of significance to whitebark 
pine establishment in the Rocky Mountains. More detailed 
observations of chipmunks in whitebark pine ecosystems 
are needed to more accurately define their role as cache 
predator and possible seed disperser. 

Golden-Mantled Ground Squirrel (Spermophilus 
lateralis) — Due to a slow foraging rate and relatively 
rare occurrence in tree canopies, the golden-mantled 
ground squirrel also consumes a limited amount of white- 
bark pine seed. This species was found in a whitebark 
pine tree only once in 188 hours of observation. It does 
collect a limited amount of seed from the ground, which 
falls from cones during foraging accidents involving other 
animals. 

Like chipmunks, groimd squirrels are known to scatter- 
hoard shallow seed caches as well as make a deeper lar- 
der hoard (MacClintock 1970; Vander Wall 1990). This 



species also begins hibernation quite early in September 
and would not have time to acquire much seed for storage. 
As with the chipmimk, few studies have been performed 
to define ground squirrel foraging and caching behavior. 

SECONDARY DISPERSERS 

Nocturnal Rodents — This group of animals, made up 
primarily of mice and voles (family Cricetidae), must also 
be considered as potential dispersal agents of whitebark 
pine seed. They were not directly observed in this study, 
but possible evidence of their foraging on whitebark pine 
seed was discovered by the shelled seed left behind on my 
simulated cache experiments. 

Smface seed caches simulating the seed that was foimd 
on the ground indicate it will not last long (Hutchins 1989; 
McCaughey and Weaver 1990). Most of the seed in these 
caches was consumed within 2 weeks after placement un- 
der trees by evidence of the seed shells left behind. Al- 
most all of the shells were left behind at the cache site, 
indicating little if any caching was done. 

There are two primary places where cricetid rodents may 
obtain whitebark pine seeds: (a) from seed that falls to the 
ground and (b) the discovery of animal (primarily nut- 
cracker) caches. A small amount of seed (-4 percent of 
the seed crop) was found to fall to the forest floor in the 
mast year of 1980. About 69 percent of seed found on the 
ground was determined imviable by examination of the 
contents inside the shell. Although these rodents may 
recache the seed, the seed was probably already placed 
in a suitable site for seedling establishment by a Clark's 
nutcracker. Thus, even if these cricetids relocate the seed 
to another cache, their positive effect on the establish- 
ment of whitebark pine is questionable at best. 

Abbott and Quink (1970), working with the winged 
eastern white pine {Pinus strobus) seeds, showed most 
caches by these rodents were made less than 15 m from 
the seed source. Thus, the habit of whitebark pine trees 
pioneering open meadows and disturbed areas is highly 
unlikely to arise from cricetid caches. Their study also 
showed that of those caches not recovered by the time the 
seed germinated, the germinated seedlings were soon con- 
sumed by these animals. This information coupled with 
the small amount of seed available on the ground for 
these rodents suggests that they could rarely be respon- 
sible for seedling establishment. Future studies need to 
look at this group of dispersal agents more closely to fur- 
ther delineate their role in whitebark pine ecosystems. 

Mice of the genus Peromyscus are members of this 
group of rodents, which is known to scatter-hoard pine 
seedhngs (Abbott and Quink 1970; Vander Wall 1992a). 
Like the chipmunks and ground squirrels, this species 
searches for seeds on the ground and will either eat them 
or cache them for later use. Because so few sound white- 
bark seeds make it to the ground, it appears the role of 
these small rodents in whitebark pine establishment is 
insignificant. 

Grizzly and Black Bears — Bears {Ursus arctos and 
U. americana) primarily obtain seed from squirrel middens 
(Kendall 1981, 1983; Mattson and Jonkel 1989), although 
black bears are known to also break branches to harvest 
the seed (Tisch 1961). Examinations of bear scat from 



168 



both species have yielded only a few intact whitebark pine 
seeds (Hutchins and Lanner 1982). Even if these seeds 
could germinate in the bear scats, they would produce an 
insignificant number of whitebark pine seedlings that are 
dispersed a very short distance fi'om the squirrel midden 
the bear raided. 

PERCENT OF SEED HARVESTED 

On forested sites, about 63 percent of the seed was har- 
vested by red squirrels and 36 percent by Clark's nut- 
cracker. The other 1 percent was harvested by all other 
animals combined (table 1). 

In contrast, on the meadow sites, which have too few 
trees for squirrels to exist, nutcrackers harvested almost 
the entire crop of whitebark pine (99 percent). Details for 
how these values were calculated are described in Hutchins 
(1982, 1990). 

In all but a year with a super-abundant cone crop, ani- 
mals harvest nearly all of the seed crop by early November 
in the Rocky Moimtains (Hutchins and Lanner 1982; 
Vander Wall 1988). By this time, no cones remain on 
the trees and seeds do not have a chance to survive long 
enough on the grotmd £ind germinate as suggested by Day 
(1967) and others. In this study I found that the seed and 
cones that have been dropped to the groimd are eaten 
within 3 weeks by various foragers. 

CONCLUSIONS 

Nutcrackers, squirrels, and the other whitebark pine 
seed forager guilds will occasionally drop some seed to the 
ground (~4 percent). Some seed is dropped intentionally 
(Vander Wall and Balda 1977) and some by accident. 
Most of this seed, however, is composed of unfilled or 
aborted ovules (69.5 percent) and could never germinate. 
Consequently, dispersal and establishment of whitebark 
pine by small rodents like deer mice, ground squirrels, 
and chipmunks would be too rare an event for a tree spe- 
cies to survive and prosper. 

We also cannot assert red squirrels to be an establisher 
of whitebark pine seedlings. As I stated in earlier work, 
I found no evidence of whitebark pine establishing on 



squirrel middens (Hutchins 1989). The midden "seed bed" 
is too deep in organic debris, and the activity on the midden 
constantly distvu-bs the soil. 

Squirrels do harvest most of the whitebark pine seed 
in the forest, and they are the major seed predator on for- 
ested sites. Their midden cone stores are also important 
for providing grizzly and black bears with an important 
fall food source (Reinhart and Mattson 1990). 

Nutcrackers, on the other hand, are one of the most im- 
portant biotic influences developing and changing subal- 
pine communities. This species alone probably accoimts 
for nearly all whitebark pine regeneration, except for 
chance happenings. As far as the bird and the tree are 
concerned, it is more profitable to cache in the open mead- 
ow. Much less predation occurs on nutcracker caches in 
the meadow (Hutchins 1990), and the small ridges are 
usually fi'ee of snow due to wind action. Consequently, 
the higher cache survival rate benefits tree regeneration 
as well as the survival of the nutcracker. This more than 
any factor may be why we see whitebark growing where 
we do — pioneering the exposed ridges, roadside cuts, 
burned sites, and meadow swales. 

The open-grown meadow trees may be the most impor- 
tant because the seeds from these trees will have a greater 
chance of being dispersed by Clark's nutcracker. These 
small tree islands lack the major seed predator (the red 
squirrel). It should be added that Reinhart and Mattson 
(1990) also foimd lower numbers of red squirrels using 
pure whitebark pine stands. These also may be good 
sources for increasing whitebark pine regeneration. 

When nutcrackers forget where they placed a cache (for 
example, Vander Wall 1982), or die, or a rodent does not 
discover the seed cache — it has a chance to germinate. 
By placing the seed in an excellent germination bed just 
below the soil surface (2-3 cm) and also hiding the seed 
fi'om easy discovery by seed predators, the bird may begin 
a new forest stand. 

Eventually, these whitebark pine trees modify the once- 
open subalpine landscape so other more shade-tolerant 
species such as Engelmann spruce and subalpine fir can 
establish themselves in this commimity (Amo 1989; 
Franklin and Dyrness 1973; Snethen 1980). The seed 
and cover produced by whitebark pine attracts a large 



Table 1 — Estimate of whitebark pine seed harvested by various animals in Wyoming 





Mean seeds 


Minutes spent 


Foraging 


Seeds 


Number of 


Percent seeds harvested 




extracted/ 


foraging/ 


days/ 


harvested/ 


individuals 


by all individuals 


Species 


minute^ 


day^ 


season 


individual 


visiting trees 


Forest 


Meadowr 


Clark's nutcracker 


7.9 


180 


91 


129,402 


448 


36.3 


99.4 


Steller's jay 


.7 


120 


55 


4,620 


11 


<.1 


.1 


Raven 


.6 


30 


53 


954 


15 


<.1 


<.1 


Noncorvids 


.7 


120 


56 


4,704 


43 


.1 


.3 


Red squirrel 


43.4 


240 


84 


874,944 


116 


63.5 




Chipmunk 


1.7 


120 


35 


7,140 


10 


<.1 


.1 



'Seasonal average from observations made during August 15 to October 11, 1980. 
^Estimate made from observed daily activity patterns. 
'Meadow area lacks squirrels. 



169 



number of vertebrates (Amo and Hoff 1989; Hutchins 
1990; Kendall and Arno 1990; Lonner and Pac 1990; 
Tomback 1978). 

SUMMARY 

Whitebark pine is dependent on animal dispersal for 
regeneration. Many animals interact with whitebark pine 
seed, but only Clark's nutcracker was found to consis- 
tently disperse seed in a way that leads to the regenera- 
tion of this pine. These long-distance dispersal agents 
should be the central focus in whitebark pine subalpine 
community management. 

ACKNOWLEDGMENTS 

I thank Steve Vander Wall for his insight on small 
mammals and pine seed dispersal. His comments were 
a great help. Susan Hutchins, Ron Lanner, and Diana 
Tomback helped with earlier versions of this manuscript. 
I thank them for their efforts. 

REFERENCES 

Abbott, H. G.; Quink, T. F. 1970. Ecology of eastern white 
pine seed caches made by small mammals. Ecology. 
51: 271-278. 

Arno, S. 1986. Whitebark pine cone crops — a diminishing 
source of wildlife food? Western Journal of Applied For- 
estry. 1(3): 92-94. 

Arno, S.; Hoff, R. J. 1989. Pinus albicaulis Engelm.: 
whitebark pine. In: Burns, R. M; Honkala, B. H., tech. 
coords. Silvics of North America. Vol. 1. Conifers. Agric. 
Handb. 654. Washington, DC: U.S. Department of Agri- 
culture, Forest Service: 269-279 

Broadbooks, H. E. 1958. Life history and ecology of the 
chipmunk {Eutamias amoenus) in eastern Washington. 
Misc. Publ. 103. Ann Arbor, MI: Museum of Zoology, 
University of Michigan: 5-42. 

Day, R. J. 1967. Whitebark pine in the Rocky Mountains 
of Alberta. Forestry Chronicle. 43: 278-282. 

Dow, D. D. 1965. The role of saliva in food storage by the 
gray jay. Auk. 82: 139-154. 

Finley, R. B., Jr. 1969. Cone caches and middens ofTam- 
iasciurus in the Rocky Mountain region. Misc. Publ. 51. 
In: Contributions in mammalogy. Lawrence, KS: Uni- 
versity of Kansas, Museum of Natural History: 233-274. 

Franklin, J. F.; Dyrness, C. T. 1973. Natural vegetation 
of Oregon and Washington. Gen. Tech. Rep. PNW-8. 
Portland, OR: U.S. Department of Agriculture, Forest 
Service, Pacific Northwest Forest and Range Experi- 
ment Station. 417 p. 

Giuntoli, M.; Mewaldt, L. R. 1978. Stomach contents 
of Clark's nutcracker in western Montana. Auk. 
95: 595-598. 

Hayashida, M. 1989. Seed dispersal by red squirrels and 
subsequent establishment of Korean pine. Forest Ecol- 
ogy and Management. 28: 115-129. 

Heller, H. C. 1971. Altitudinal zonation of chipmunks 
{Eutamias): interspecific aggression. Ecology. 
52(2): 312-319. 



Hutchins, H. E. 1982. The role of birds and mammals 
in the dispersal and establishment of whitebark pine. 
Logan, UT: Utah State University. 67 p. Thesis. 

Hutchins, H. E. 1990. Whitebark pine seed dispersal and 
establishment: who's responsible? In: Schmidt, W. C; 
McDonald, K. J., comps. Proceedings — symposiimi on 
whitebark pine ecosystems: ecology and management of 
a high-mountain resource; 1989 March 29-31; Bozeman, 
MT. Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Depart- 
ment of Agriculture, Forest Service, Intermountain Re- 
search Station: 245-255. 

Hutchins, H. E.; Lanner, R. M. 1982. The central role of 
Clark's nutcracker in the dispersal and establishment 
of whitebark pine. Oecologia. 55: 192-201. 

Kendall, K. C. 1981. Bear use of pine nuts. Bozeman, MT: 
Montana State University. 27 p. Thesis. 

Kendall, K. C. 1983. Use of pine nuts by grizzly and black 
bears in the Yellowstone area. Interagency conference 
on bear research and management. In: Bears — their 
biology and management. Calgary, AB: International 
Association for Bear Management. 5: 166-173. 

Kendall, K. C; Arno, S. F. 1990. Whitebark pine— an im- 
portant but endangered wildlife resource. In: Schmidt, 
W. C; McDonald, K. J., comps. Proceedings — symposium 
on whitebark pine ecosystems: ecology and manage- 
ment of a high-mountain resource; 1989 March 29-31; 
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Research Station: 264-273. 

Lanner, R. M. 1980. Avian seed dispersal in the ecology 
and evolution of limber and whitebark pines. In: Sixth 
North American forest biology workshop proceedings. 
Edmonton, AB: University of Alberta: 14-48. 

Lanner, R. M. 1982. Adaptations of whitebark pine for 
seed dispersal by Clark's nutcracker. Canadian Journal 
of Forest Research. 12(2): 391-402. 

Lanner, R. M.; Vander Wall, S. B. 1980. Dispersal of lim- 
ber pine seed by Clark's nutcracker. Joiirnal of For- 
estry. 78(10): 637-639. 

Lonner, T. N.; Pac, D. F. 1990. Elk and deer use of white- 
bark pine forests in southwestern Montana: an ecologi- 
cal perspective. In: Schmidt, W. C; McDonald, K. J., 
comps. Proceedings — symposium on whitebark pine eco- 
systems: ecology and management of a high-mountain 
resource; 1989 March 29-31; Bozeman, MT. Gen. Tech. 
Rep. INT-270. Ogden, UT: U.S. Department of Agricul- 
ture, Forest Service, Intermountain Research Station: 
237-244. 

MacClintock, D. 1970. Squirrels of North America. New 
York: Van Nostrand Reinhold. 184 p. 

McCaughey, W. W.; Weaver, T. 1990. Biotic and micro- 
site factors affecting whitebark pine establishment. In: 
Schmidt, W. C; McDonald, K. J., comps. Proceedings — 
symposivun on whitebark pine ecosystems: ecology and 
management of a high-mountain resource; 1989 
March 29-31; Bozeman, MT. Gen. Tech. Rep. INT-270. 
Ogden, UT: U.S. Department of Agriculture, Forest Ser- 
vice, Intermountain Research Station: 140-150. 

Reinhart, D. P.; Mattson, D. 1990. Red squirrels in the 
whitebark zone. In: Schmidt, W. C; McDonald, K. J., 



170 



comps. Proceedings — symposium on whitebark pine eco- 
systems: ecology and management of a high-mountain 
resource; 1989 March 29-31; Bozeman, MT. Gen. Tech. 
Rep. INT-270. Ogden, UT: U.S. Department of Agricul- 
ture, Forest Service, Intermountain Research Station: 
256-263. 

Smith, C. C. 1968. The adaptive nature of social organiza- 
tion in the genus of tree squirrels Tamiasciurus. Eco- 
logical Monographs. 38: 31-63. 

Smith, C. C. 1970. The coevolution of pine squirrels 
{Tamiasciurus) and conifers. Ecological Monographs. 
40: 349-371. 

Smith, C. C; Balda, R. P. 1979. Competition among in- 
sects, birds, and mammals for conifer seed. American 

Zoologist. 19: 1065-1083. 
Snethen, K. L. 1980. Whitebark pine {Pinus alhicaulis) 

invasion of a subalpine meadow. Logan, UT: Utah State 

University. 76 p. Thesis. 
Tevis, L., Jr. 1953. Stomach contents of chipmunks and 

mantled ground squirrels in northeastern California. 

Journal of Mammalogy. 34: 316-324. 
Tisch, E. L. 1961. Seasonal food habits of the black bear 

in the Whitefish Range of Northwestern Montana. 

Missoula, MT: University of Montana. 67 p. Thesis. 
Tomback, Diana F. 1978. Foraging strategies of Clark's 

nutcracker. The Living Bird. 16: 123-161. 
Tomback, Diana F. 1982. Dispersal of whitebark pine 

seeds by Clark's nutcracker: a mutualism hypothesis. 

Journal of Animal Ecology. 51: 451-467. 



Turcek, F. J.; Kelso, L. 1968. Ecological aspects of food 
transportation and storage in the Corvidae. Commimi- 
cations in Behavioral Biology. 1: 277-297. 

Vander Wall, S. B. 1982. An experimental analysis of 
cache recovery in Clark's nutcracker. Animal Behaviour. 
30: 84-94. 

Vander Wall, S. B. 1988. Foraging by Clark's nutcracker 
on a rapidly changing pine seed source. Condor. 90: 
621-631. 

Vander Wall, S. B. 1990. Food hoarding in animals. 
Chicago, IL: University of Chicago Press. 445 p. 

Vander Wall, S. B. 1992a. Establishment of Jeffrey pine 
seedings from animal caches. Western Jovirnal of Ap- 
plied Forestry. 7(1): 14-20. 

Vander Wall, S. B. 1992b. [Personal communication]. 
July. Reno, NV: University of Nevada, Department 
of Biology. 

Vander Wall, Stephen B.; Balda, R. P. 1977. Coadapta- 
tions of Clark's nutcracker and the pinon pine for effi- 
cient seed harvest and dispersal. Ecological Monographs. 
47: 89-111. 

Vander Wall, Stephen B.; Balda R. P. 1981. Ecology 
and evolution of food-storage behavior in conifer-seed 
caching corvids. Zeitschrift fiir Tierpsychologie. 56: 
217-242. 

Vander Wall, Stephen B.; Hutchins, Harry E. 1983. 
Dependence of Clark's nutcracker, Nucifraga Colum- 
biana, on conifer seeds during the postfledging period. 
The Canadian Field Naturahst. 97(2): 208-214. 



171 



SIZE OF PINE AREAS IN RELATION 
TO SEED DISPERSAL 



Hermann Mattes 



Abstract — Pine seeds are either wind- or bird- (animal-) 
dispersed. Only well-adapted seeds support large ranges, which 
are of similar size in both types of dispersal. Seeds adapted to 
wind dispersal should have a ratio of seed weight to wing length 
below 1.8 mg/mm. Bird-dispersed seeds have a weight between 
90 and 550 mg. Species with seeds above 100 mg but without 
zoochorous features have small or even relictic ranges. Large 
seeds are advantageous under severe conditions and in competi- 
tion. Consequently, the bird-pine mutualism contributes to 
larger ranges of those pine species. 



Distribution ranges of plant and animal species are de- 
termined by four main factors: ecological potency of the 
species; effectiveness of seed dispersal; size of area avail- 
able; and time available. Ecological studies on plants and 
vegetation mostly focus on site conditions. Indeed, compe- 
tition for nutrients, water, light, and so on most often 
turn out to be the main factor limiting distribution. How- 
ever, first of all, seeds must be able to reach the site in 
question. Therefore, seed dispersal is an important factor 
in plant ecology. Patchy and irregular patterns of species 
on ruderal or fallow ground are a result of dissemination. 
Otherwise vegetation history in the Holocene gives evi- 
dence that time since last glaciation was not long enough 
for full expansion of all species. We expect effectiveness 
of seed dispersal to be correlated with rapidness of expan- 
sion and with size of range. 

Small, winged seeds are easily disseminated by wind, 
which is present almost everywhere and emytime. Com- 
petition of the seedling within a closed vegetation cover 
as well as harsh environmental conditions require large 
seeds with a high amount of nutrient reserves. Many 
pine species have developed an almost obligate mutualis- 
tic relationship to birds (Nucifraga, Gymnorhinus) for 
seed dispersal. Pines with bird-dispersed, large seeds be- 
came dominating tree species at arctic and alpine timber 
lines, in xeric environments as well as in mesic lowland 
forests. 

The genus Pinus with more than 100 species provides 
a good chance to test effectiveness of seed dispersal. All 
pine seeds are of a rather similar structure; the most obvi- 
ous differences are seed size and wing length. The pri- 
mary type is very likely an anemochorous seed with a 
long wing. Second, many species have developed large 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Hermann Mattes is Professor, Institut fur Geographie, Abteilung 
Landschaftsokologie, Westfalische Wilhelms-Universitat, Robert-Koch- 
Str. 26, W-4400 MOnster, Germany. 



seeds with shortened or missing vnngs. These seeds are 
disseminated mainly by birds, or occasionally by squirrels. 
Apart from that, seeds and cones show many morphologi- 
cal and phenological adaptations for dispersal, which are 
discussed elsewhere (Lanner 1980, 1982; Vander Wsdl and 
Balda 1977). 

In this paper some relations of seed characteristics and 
size of ranges are discussed. The main hjrpothesis is to 
find the larger ranges within pine species well adapted 
to seed dispersal. 

MATERIALS AND METHODS 

From 65 out of about 106 pine species in total enough 
detailed information is available for a comparison of dis- 
persal abilities and range sizes. The pine species are 
numbered following Mirov (1967). 

Distribution ranges are of a complex nature. Effective- 
ness of dispersal would be measured best by the propor- 
tion of the potential area that really has been occupied by 
the species. Unfortunately, we do not know exactly the 
potential ranges of almost all plant species. An interest- 
ing idea would be to measure the distances or Eireas occu- 
pied after the end of the last glaciation. However, refugials 
and paths of dispersal are very incompletely known. 
What we can clearly see are area and distances within 
ranges. However, it seems not to be adequate to use the 
area of a range. It is highly influenced by size of the re- 
gion (the continent or vegetation zone that is inhabited 
by the species concerned). Error should be less using the 
distance between the outermost points of a range. This 
was calculated as the orthodrome distance. 

Split or disjunct ranges have been treated in the same 
way as continuous ranges. It has proved to be too difficult 
to decide whether an interrupted range was due to natu- 
ral factors or to hxmian influences, especially in regions 
that were densely populated for a long time such as East 
Asia and the Mediterranean. 

In the present study all 105 pine species listed in Mirov 
(1967), and in addition Pinas longaeva, are considered. 
Subspecies have not been considered. Reinge maps £ind 
information was taken from Kriissmann (1968), Little 
(1971), Meusel and others (1965), and Mirov (1967). 

In only a few experiments flying ability of pine seeds 
was examined (Lanner 1985; Muller-Schneider 1977). 
To estimate dispersal abilities of numerous species, indi- 
rect methods have been used. Simple, but reliable meas- 
lu-ements for anemochorous seeds are weight of cleaned 
seeds and wing length. Such data are available for most 
of the species (Kriissmann 1972; USDA 1974). The ratio 
of seed weight to wing length is used as an indicator for 
potency of dispersal. Volume and surface of the seed and 
shape and surface of the vring would be of great interest, 



172 



but these data are not available for most species. Zooch- 
orous species are even more difficult to evaluate. Seed 
and cone characteristics of species fully developed for bird 
dispersal are compared morphologically with less-adapted 
species. 

DISPERSAL OF PINE SEEDS 

Seed dispersal either by wind or by birds has caused 
specific adaptions. Most of the pine species can be attrib- 
uted to these two main types. A narrow seed-weight to 
wing-length ratio is considered to be favorable to flight 
ability. From figure 1 with 46 pine species Usted we can 
assume: 

• In general, the length of seed wings increases until 
seed weight reaches about 90 mg (P. palustris: 93 mg/ 
36 mm); however, the ratio of seed weight to wing length 
increases obviously with increasing seed weight. 

• Winged seeds of more than about 100 mg have an 
unfavorably high ratio of seed weight to wing length; wing 
length is insvifficient even in species with huge cones and 
long cone scales. These seeds are expected to have no 
adequate flight abilities. 

• The smallest seed weight of a wingless seed species 
{P. flexilis) is 93 mg; most of the species with heavier 
seeds are wingless, or seed wings separate easily (P. penta- 
phylla No. 83, P. himekomatsu No. 84, P. strobiformis 
No. 35; also in P. pinea No. 73). 

Comparing morphological features with area size we 
can conclude the following (fig. 2, table 1). 

• Wind-dispersed pine species with a low ratio of seed 
weight to wing length often have larger distribution 
ranges than species with a high ratio. 

• There is evidence for a log-linear relationship for the 
largest ranges of each class of seed weight/wing length ra- 
tio. Species with a ratio higher than 2 mg/mm are less 
distributed on an average. It is supposed that dispersal 
abilities limit the range of distribution. A remarkable ex- 
ception is P. pinea (No. 73). Its origin is not known, and 
it has been cultivated in many places since Phoenicean 
times, approximately 4,000 years ago. Therefore, the 
whole present range had to be taken into consideration. 

• Many species do not reach an area as large as could 
be expected in view of their theoretical abilities for seed 
dispersal. This can be partly explained by geographical 
and ecological reasons. Species of southeastern North 
America (Nos. 22-31) depend on a relatively small area 
limited by the ocean or woodless plains. The range of 
Cahfomian pines (Nos. 5, 6, 17-19) of the Mediterranean 
type is limited by climate. Mexican pines (seed data only 
for P. chihuahua and patula, Nos. 40 and 61) also are re- 
stricted geographically and ecologically to narrow moxm- 
tainous belts. Pinus peuce and P. heldreichii (Nos. 71 and 
75) are endemic to the Balkan mountains in southeastern 
Europe. 

• Ranges of bird-dispersed pines are of similar size as 
those of wind-dispersed pines; differences are mainly due 
to the very restricted range of the bird-dispersed P. quad- 
rifolia in America and the huge range of wind-dispersed 
P. sylvestris in Eurasia. Some bird-dispersed pine ranges 
are among the largest of all pines. Concerning size of 



ranges, dispersal by birds (and sqmrrels) is as efficient 
as dispersal by wind. 

• Pine ranges in America are smaller than those in 
Eurasia. This is an effect of the larger land mass in the 
Old World hemisphere. Only species with well-adapted 
dispersal mechanisms are affected; species with relicted 
ranges show no difference in area size. 

GENUS PINUS DISTRIBUTION 

Most of the pine species are holarctic. Pines occupy 
a wide variety of sites of which the most extreme are 
located at timber lines in cold and arid climates as well 
as in very dry or perhvunid regions in the subtropics and 
tropics. Many of the pine species are restricted to specific 
sites. Thus, an effective dispersal agent is needed to 
reach suitable sites. 

Recent centers of species diversity (fig. 3) are Mexico 
and the southwestern States of the U.S.A. Secondary cen- 
ters of diversity are eastern North America, eastern Asia, 
and the Mediterranean region. These were refugial areas 
during glaciation. Although none of these areas is neces- 
sarily the origin place of the genus Pinus, recent regions 
of high species numbers of pines are assumed to be places 
where pines have existed for a long time. 

In contrast, there are large areas populated only by a 
few pine species. This is the case especially in the present 
boreal zone. Also, we find relatively few pine species in 
Central America south of Tehuantepec, on the Caribbean 
Islands, in the Himalayas, and in southeastern Asia. 
Probably, these areas have been occupied by pines for 
only a relatively short time. 

The boreal zone was reoccupied by pines in the postgla- 
cial time, and we know that some species are stiU expand- 
ing (for example see Gorchakovsky 1993, for Pinus sibi- 
rica). Nevertheless, their ranges belong to the largest 
ranges of pine. All boreal pine species are among the spe- 
cies best adapted either for wind dispersal (figs. 1 and 2; 
Nos. 16, 20, 21, 32, and 69) or bird dispersal (Nos. 67 and 
68). Wind-dispersed pines have excellent seed-weight/ 
wing-length ratios from 0.4 to 1.1 mg/mm. 

Data are available for only a few species of southern 
marginal regions (P. wallichiana, No. 91; P. merkusii, 
No. 101; P. khasya, No. 104). With a ratio of seed weight/ 
wing length of 1.2 to 1.7 mg/mm, adaptation for wind dis- 
persal in these species is relatively good. 

Species that were more widely distributed in the Ter- 
tiary or early Pleistocene, and that now occupy relictic 
ranges only, show very different ratios of seed weight/ 
wing length. For P. aristata (No. 5; 1.2 mg/mm) and per- 
haps P. balfouriana (No. 6; 2.1 mg/mm) and P. longaeva 
(No. 6a) seed dispersal might not be a restricting factor. 
This svirely is the case in P. torreyana (No. 11). 

Species with reduced seed wings have less chances to 
be dispersed by wind. Some of them are on the way to 
developing zoochorous features. An advanced state of 
zoochory shows in especially P. strobiformis (No. 35), and 
somewhat less in P. pentaphylla (No. 83) and P. himeko- 
matsu (No. 84) (Hayashida 1989); they have large seeds 
and wings that easily separate fi-om the seed. These spe- 
cies are already mainly disseminated by birds and ani- 
mals. Pinus peuce (No. 71) has a small seed (40 mg) and 



173 



length of seed wing 



0.4 
0.4 

0.5 n 

0.4 ] 
0.4 D 
0.7 
0.6 
0.S 
0.6 h 
0.7] 
1.6 3 

1.13 
1.1 J 



0.7 1 
1.5 □ 



16 
5.9 

n 84 



seed weight 

banks iana 32 
contorta 16 
sylvestris 69 
clausa 28 
serotina 30 
pa tula 61 
resinosa 21 
virginiana 27 
chihuahua 40 
pungens 31 
halepensis 77 
strobus 20 
nigra 74 

1.0 p monticola 4 
attenuata 17 
heldreichii 75 

1.3 p merkusii 101 

2.1 □ aristata 6 

1.3 ZJ elliottii 23 
1.5 tzi radiata 19 

ponderosa 14 
peuce 71 
engelmannii 
pinaster 80 
wall ichi ana 

□ palustris 

□ flexilis 2 

□ pumila 6 8 
m! pentaphylla 83 
~i jeffreyi 13 

:l strobiformis 35 
I albicaulis 1 
1 lamber tiana 3 



1.3 
5.7 



1.6 i 

1.9 ZD 



1.7 
2.6 



7.7 



13 



[j378 
56 



53 

91 
22 



181 



cembra 7 0 
edulis 8 
j sibixica 67 

U armandii 86 
i coulteri 12 

. 1 monophylla 7 

1 gerardiana 92 
" I cembroides 36 
• ^ I quadrifolia 9 

: "~1 koraiensis 82 

I pinea 7 3 
: . ' I sabiniana 

torievana] ll 



10 



[mml 30 20 



10 



0 0 100 200 300 400 500 600 700 800 900 [mgl 



Figure 1— Wing length (mm, left), weight (mg, right), and ratio of wing length to weight (mm/mg, 
middle) of seeds of 46 pine species. Numbers following species names according to Mirov (1967). 
For sources see text. Species are arranged according to seed weight. 



is not very attractive for nutcrackers, although it is used 
for hoarded food sometimes as is P. longaeva (Lanner 
1988). It is assumed that dispersal by nutcrackers is of 
some relevance in establishing these two species at tim- 
berline. Ranges of both species are small. 



Other pine species such as P. lambertiana, P. sabiniana, 
P. torreyana, P. coulteri, and P. jeffreyi (Nos. 3, and 10-13) 
have large seeds but lack further adaptations for disper- 
sal by birds and animals. Since these species have no ef- 
fective dispersing agent, their ranges are relatively small. 



174 



{mg/mmj in 
403 6-j 



^ 148 
"5 

55 
5 7.4 

f 

2.7 
1.0 



5 -1 
4 
3 - 
2 H 

1 
0 



0.4 -1 -i 
0.6 -1.5 



11 



71 



•63 



7 36 1 86 
ftX'n ** **** • ■ * 

92 8 70 35 2 8 2 6 7 6 8 



' 73 



84 



• 83 



'12 



• 3 

• 13 



• Ca 



. 53 



-91 



• 78 
•22 

5 '19, 24';i-0l'77 
23 1 a . • • •14 

40.,7.95 'ZS^^ 

.28 26 ^52^ .32 
•l6 



•69 



5.5 
246 



6.0 
404 



6.5 
685 



7.0 7.5 8.0 8.5 9.0 9.5 



In 



1097 1808 2961 4815 8103 13360 [km] 
size of range 



Figure 2 — Size of ranges in relation to the ratio wing length/ 
seed weight of 44 wind-dispersed pines; for comparison, 
range size of 13 bird-dispersed pine species on top of the 
figure. For code numbers for pine species see appendix; 
for sources see text. 



PiniLs pinea, as mentioned earlier, is an exception because 
it has been cultivated throughout history. 

CONCLUSIONS 

Competition in seedlings causes a high selective pres- 
sxire for large seeds. Even among wind-dispersed pines 
seed size tends to enlarge despite a strong negative effect 
on dispersal ability. Obviously, at 90 to 100 mg seed 
weight dispersal success by wind has diminished so far 
that other agents are essential. From 24 haploxylon pine 
species, 19 species have wingless or almost wingless and 



large seeds, which are proven to be or expected to be bird 
dispersed. There is no doubt that zoochory has developed 
in parallel in several sections of haploxylon pines (Lanner 
1989; Tomback and Linhart 1990). Three diploxylon spe- 
cies (P. torreyana, No. 11; P. sabiniana, No. 12; P. pinea, 
No. 73) have large, nearly wingless seeds but lack further 
adaptations for zoochory. 

Range sizes of bird-dispersed pines are not obviously 
different from those of wind-dispersed pines. Dissemina- 
tion of pine seeds by birds is at least as effective as wind 
dispersal. The interaction of pines and nutcrackers or 
jays is well balanced. Even scattered stands of zoochorous 



Table 1 — Average size of ranges (km distance) of pine species with different dispersal agents. Wind-dispersed 
pines are arranged according to their seed characteristics (seed weight/wing length in mg/mm) 



Distribution ranges 

Pine species dispersed by: America Eurasia Global 



Birds 


1,557 


in 


= 7) 


2,680 


in = 


6) 


2,075 


in 


= 13) 


Wind (0.4-0.8 mg/mm) 


1,913 


in 


= 13) 


M,247 


in = 


3) 


2,351 


in 


= 16) 


Wind (0.9-2.0 mg/mm) 


1,926 


in 


= 9) 


2,610 


in = 


7) 


2,225 


in 


= 16) 


Wind (>2.0 mg/mm) 


1,121 


in 


= 7) 


21,055 


in = 


4) 


21,097 


in 


= 11) 


Species considered above 


1,693 


in 


= 36) 


22,566 


in = 


20) 


21,994 


in 


= 57) 


All species 


1,413 


in 


= 67) 


1,779 


in= 


39) 


1,548 


in 


= 106) 



'Without P. sylvestris, average range size is 2,150 km. 
^Numbers without P. pinea. 



175 




176 



pine trees support all needs for territories of nutcrackers 
concerning breeding, feeding, overwintering, and social 
life. Nutcrackers and jays have still preserved character- 
istic euryoecious features of corvids, especially omnivory. 
In Eurasia, where there is almost no other animal com- 
peting with the nutcracker for seed hoarding, the range 
of the nutcracker surpasses the ranges of zoochorous 
pines in many regions. 

Within wind-dispersed pines we find that large-sized 
ranges suppose small seeds with a ratio of weight to wing 
length below approximately 1.8 mg/mm, but not vice- 
versa because of geographical barriers or specialized fea- 
tures of a species. However, pine species with a very high 
ratios of seed weight/wing length (above 1.8 mg/mm) al- 
ways have restricted areas. 

Altogether, pine seed dispersal by birds was very suc- 
cessful dvuing evolution. Pine species became able to oc- 
cupy additional and more different sites, as well as under 
extreme conditions at the borders of tree growth as under 
mesic conditions. Consequently, bird-dispersed pines 
were able to enlarge their ranges. 

REFERENCES 

Critchfield, William B.; Little, Elbert L., Jr. 1966. Geo- 
graphic distribution of the pines of the world. Misc. 
Pub. 991. Washington, DC: U.S. Department of Agricul- 
ture, Forest Service. 97 p. 

Gorchakovsky, P. L. 1993. Distribution and ecology of Si- 
berian stone pine (Pinus sibirica) in the Urals. (These 
proceedings). 

Hayashida, Mitsuhiro. 1989. Seed dispersal and regenera- 
tion patterns of Pinus parviflora var. pentaphylla on 
Mt. Apoi in Hokkaido. Research Bulletin of the College 
Experiment Forests, Faculty of Agriculture, Hokkaido 
University. 46(1): 177-190. 

Kriissmann, Gerd. 1968. Die Baume Europas. Berlin and 
Hamburg: Parey Verlag. 140 p. 

Kriissmann, Gerd. 1972. Handbuch der Nadelgeholze. 
Hambiu-g und Berlin: Parey- Verlag. 

Lanner, Ronald M. 1980. Avian seed dispersal as a factor 
in the ecology and evolution of limber and whitebark 



pines. Proc. 6th North American Forest Biology Work- 
shop, University of Alberta, Edmonton, Canada: 15-48. 

Lanner, Ronald M. 1982. Adaptations of whitebark pine 
for seed dispersal by Clark's nutcracker. Canadian 
Journal of Forest Research. 12(2): 391-402. 

Lanner, Ronald M. 1985. Effectiveness of the seed wing of 
Pinus flexilis in wind dispersal. Great Basin Naturalist. 
45(2): 318-320. 

Lanner, Ronald. 1988. Dependence of Great Basin bristle- 
cone pine on Clark's nutcracker for regeneration at high 
elevation. Arctic and Alpine Research. 20(3): 358-362. 

Lanner, Ronald M. 1989. Biology, taxonomy, evolution, 
and geography of stone pines of the world. In: Schmidt, 
Wyman C; McDonald, Kathy J., comps. Proceedings — 
symposium on vv^hitebark pine ecosystems: ecology and 
management of a high-mountain resource. Gen. Tech. 
Rep. INT-270. Ogden, UT: U.S. Department of Agricul- 
ture, Forest Service, Intermountain Research Station: 
14-24. 

Little, Elbert L., Jr. 1971. Atlas of United States trees. 

Vol. 1: Conifers and important hardwoods. Misc. Publ. 

1146. Washington, DC: U.S. Department of Agriculture. 

9 p. and 200 maps. 
Meusel, H.; Weinert, E.; Jager, E. 1965. Vergleichende 

Chorologie der zentraleuropaischen Flora. Bd. I. Jena: 

Fischer- Verlag. 9 p. and 200 maps. 
Mirov, Nicholas T. 1967. The genus Pinus. New York: 

Ronald Press. 602 p. 
Miiller-Schneider, Paul. 1977. Verbreitungsbiologie 

(Diasporologie) der Blutenpflanzen. Veroffentlichungen 

des geobotanischen Institutes der ETH Zurich, Stiftung 

Riibel. Vol. 61. 226 p. 
Tomback, Diana F.; Linhart, Yan B. 1990. The evolution 

of bird-dispersed pines. Evolutionary Ecology. 4: 185-219. 
U.S. Department of Agriculture, Forest Service. 1974. 

Seeds of woody plants in the United States. Agric. 

Handb. 450. Washington, DC. 838 p. 
Vander Wall, Stephen B.; Balda, Russ P. 1977. Coadapta- 

tions of the Clark's nutcracker and the pinyon pine for 

efficient harvest and dispersal. Ecological Monographs. 

Durham. 47(1): 27-37. 



177 



APPENDIX: CODING LIST OF NUMBERS OF PINE SPECIES CONSIDERED 
(AFTER MIROV 1967) wi^c»ii^r.rtr.u 



Western North America 

Haploxylon pines 

1 albicaulis 

2 flexilis 

3 lambertiana 

4 monticola 

5 balfouriana 

6 aristata 
6a longaeva 

7 monopylla 

8 edulis 

9 quadrifolia 

Diploxylon pines 

10 sabiniana 

11 torreyana 

12 coulteri 

13 jeffreyi 

14 ponderosa 

16 contorta 

17 attenuata 

18 muricata 

19 radiata 

Eastern North America 

Haploxylon pines 

20 strobus 

Diploxylon pines 

21 resinosa 

22 palustris 

23 elliotti 

24 toeda 

25 echinata 

26 glabra 

27 virginiana 

28 clausa 

29 rigida 

30 serotina 

31 pungens 

32 banksiana 



Mexico 

Haploxylon pines 

35 strobiformis 

36 cembroides 

Diploxylon pines 
40 chihuahua 
53 engelmanni 

Northern Eurasia 

Haploxylon pines 

67 sibirica 

68 pumila 

Diploxylon pines 

69 sylvestris 

Mediterranean Region 

Haploxylon pines 

70 cembra 

71 peuce 

Diploxylon pines 

73 pinea 

74 wi^m 

75 heldreichii 

76 montana i=mugo) 

77 halepensis 

78 brutia 
80 pinaster 

Eastern Asia 

Haploxylon pines 

82 koraiensis 

83 pentaphylla 

84 himekumatsu 
86 armandi 

91 wallichiana (=griffithi) 

92 gerardiana 

Diploxylon pines 
95 densiflora 
101 merkusii 
104 khasya 



178 



THE REGENERATION PROCESS 
OF WHITEBARK PINE 

Ward W. McCaughey 



Abstract — ^Whitebark pine (Pinus albicaulis Engelm.) regener- 
ates similarly to European and Asian stone pines. Our knowl- 
edge of this process is essential because whitebark pine is impor- 
tant to the North American grizzly bear {Ursus arctos horribilis), 
other wildlife species, hydrology of high-elevation ecosystems, 
and esthetics. This paper simimarizes available information on 
the whitebark pine regeneration process, beginning from bud ini- 
tiation through germination and seedling survival. Major factors 
limiting germination and regeneration success are discussed. We 
continue to gain knowledge about the regeneration process of 
whitebark pine, but further research is needed to fully imder- 
stand delayed germination mechanisms and habitat require- 
ments for optimimi regeneration success. 



Whitebark pine (Pinus albicaulis Engelm.) is an impor- 
tant food source for the endangered grizzly bear (Ursus 
arctos horribilis), red squirrels (Tamiasciurus hudsonicus), 
the Clark's nutcracker (Nucifraga columbiana Wilson), 
and a multitude of other birds and mammals (Hutchins 
and Lanner 1982; Kendall 1983; Knight and others 1987). 
It is also significant as a hydrologic stabilization plant 
and an esthetic feature of high-elevation communities 
(Schmidt and McDonald 1990). It provides important 
cover for wildlife and is used as an ornamental for land- 
scaping but has only minor significance as a timber pro- 
ducing species (McCaughey and Schmidt 1990). 

General information is available and is summarized 
in this paper on the natiu-al regeneration process of white- 
bark pine, but more specific facts are needed for efficient 
management of this North American stone pine (Amo and 
Hoff 1989; McCaughey and Schmidt 1990). This paper 
describes the regeneration process beginning firom bud 
initiation throiigh germination and seedling survival. 

CONE AND SEED DEVELOPMENT 

Whitebark pine cone and seed development begins vnth 
cone initiation and ends with seed matiiration. Climatic 
conditions influence the timing of development and matu- 
ration of cone initiation and seed maturation. 

Cone Initiation 

The process of cone initiation and development of stami- 
nate and ovulate strobili for whitebark pine has not been 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Ward W. McCaughey is a Research Forester, Intermountain Research 
Station, Forest Service, U.S. Department of Agriculture, Bozeman, MT 
59717-0278. 



specifically studied. Mirov (1967) and Krugman and 
Jenkinson (1974) provided generalized descriptions of the 
initiation process for the genus Pinus. Female and male 
cone initiation of whitebark pine probably occiu- during 
or just prior to winter bud formation from mid-August to 
mid-September (Schmidt and Lotan 1980). 

Cone Development 

Female and male buds overwinter and begin further 
development and growth in April and May depending on 
elevation. In 1992, staminate cones matured and shed 
pollen during Jxme and early Jvdy at high elevations 
(2,500-2,600 m) in southwestern Montana (McCaughey 
1992) and in May and June at lower elevations (1,550- 
1,650 m) in northern Idaho (Hoff 1992). Mature male 
cones are about 1 cm wide by 1 cm long and female cones 
are 1 cm wide by 2 cm long during early development 
when they are pollen receptive. After pollination female 
cones grow to a first-year size of about 3 cm long by 2 cm 
wide. Female cones remain on the tree, while male cones 
fall off after pollen dispersal. Pollination occurs from late 
Jime to late Jvdy, and fertilization occxirs about 12 to 13 
months later, simultaneous with rapid cone enlargement 
during June and July (Krugman and Jenkinson 1974). 

Seed Maturation 

Following fertilization the embryo develops and dif- 
ferentiates into cotyledons, plumule, and radicle (Mirov 
1967). Whitebark seeds reach maturity between mid- 
August and mid-September depending on elevation and 
climatic conditions during the ripening process (Krugman 
and Jenkinson 1974; McCaughey, in press a). The entire 
process from cone initiation to cone and seed matiuity 
takes about 24 months. 

Cone and Seed Insects and Diseases 

Whitebark pine cones and seeds are exposed to insects 
and disease that reduce cone and seed siirvival during the 
2 years of development (Edwards 1990). Cone and seed 
insects that damage whitebark pine are cone worms (Dio- 
ryctria spp. and Eucosma spp.), cone beetles (Conophtho- 
rus spp.) (Bartos and Gibson 1990), and midges and seed 
chalcids (Megastigmus spp.) (Dewey 1989). Siroccocus 
strobilinus Preuss is a seed-borne disease that kills white- 
bark pine seedlings in niu-series and in natxu-al stands. 
Calocypha fulgens (Pers.) Boud. (anamorph = Geniculcden- 
dron pyriforme Salt), referred to as a seed or cold fungus, 
may cause preemergence seed loss (Hoff and Hagle 1990). 
Pre- and postemergence damping-off diseases (Fusarium 
spp.) may cause extensive mortality, especially in slowly 
emerging seedlings (Landis and others 1990). 



179 



SEED DISPERSAL 

Whitebark pine cones are well suited to the Clark's nut- 
cracker, a forest bird that is the major disperser of white- 
bark pine seeds (Arno and Hoff 1989; Hutchins and 
Lanner 1982; Lanner 1980; Lanner and Vander Wall 1980; 
McCaughey and Schmidt 1990; Tomback 1982). 

Cone Attributes 

Unlike other conifers such as western white pine (Pinus 
monticola Dougl.) and limber pine {Pinus flexilis James), 
whose cone scales fully open to release seed, whitebark 
pine cone scales only open partly. The scale base does not 
break away from the cone axis; the wingless seeds are 
held firmly in place yet fully exposed and easily accessible 
to nutcracker extraction (Eggers 1986; Hutchins and 
Lanner 1982). Nutcrackers have strong, pointed bills like 
other species of the family Corvidae. Upper portions of 
the cone scales are easily broken by nutcrackers along 
a thin fracture zone where the massive apophysis tapers 
to a thin cross section beneath the seed-bearing cavities 
(Lanner 1982). A nutcracker can store over 100 white- 
bark pine seeds in its sublingual pouch, a saclike modifi- 
cation on the floor of the mouth (Bock and others 1973). 

Nutcracker Caching 

Whitebark pine regeneration depends almost entirely 
on nutcracker seed selection and caching habits. A nut- 
cracker appears to discriminate between good, aborted, 
insect-infected, or diseased seeds by rattling each seed 
in its bill before depositing it in the sublingual pouch 
(Hutchins and Lanner 1982; Tomback 1978; Vander Wall 
and Balda 1977). Whitebark pine seeds are dispersed by 
nutcrackers up to 22 km from their source. Caches are 
buried 2 to 4 cm deep with 1 to 25 or more seeds per cache 
(fig. 1) (Tomback 1978; Vander Wall and Balda 1977). 
One nutcracker can store an estimated 22,000 to 98,000 
whitebark pine seeds each year when seed is available 
(Hutchins and Lanner 1982; Tomback 1978; Vander Wall 
and Balda 1977). Estimates of food requirements indicate 
that nutcrackers may store three to five times as many 
seeds as needed (Tomback 1983). 

Secondary Dispersers 

Nutcrackers compete with other seed consumers for 
whitebark pine seeds. Red squirrels harvest cones and 
seeds from mid-July to early November, greatly reducing 
the availability of seed for nutcrackers (Eggers 1986; 
Hutchins and Lanner 1982; Reinhart and Mattson 1990; 
Smith 1968). Squirrels store cones in middens and seeds 
in caches on the forest floor (Hutchins and Lanner 1982; 
Reinhart and Mattson 1990). 

Grizzlies obtain whitebark pine seeds primarily from 
squirrel middens (Kendall 1983). Germination probabil- 
ity from squirrel-cached cones and seed is low due to their 
deep caching habits (>7 cm) and small numbers of midden 
sites. H5^ocotyl growth of whitebark pine is only 3 to 4 cm; 
emergence of cotyledons above the soil surface in a 




Figure 1 — Nutcracker cache site with five white- 
bark pine germinants growing on a litter seedbed 
near a log. 



squirrel midden is unlikely (Lanner 1982). Squirrels 
feed on the cones and seed during the winter and spring 
months thus constantly disturbing the middens (Lanner 
1982). 

Many other animals feed on whitebark pine seed and are 
considered secondary dispersers because of the low prob- 
ability of germination following their feeding and caching 
(Hutchins and Lanner 1982; McCaughey and Schmidt 
1990). Other common animals that harvest whitebark 
pine seed either from the cones directly or indirectly from 
the ground or other animal caches are: birds — ^William's 
sapsucker (Sphyrapicus thyroideus), hairy woodpecker 
{Picoides villosus), white-headed woodpecker {P. albolar- 
vatus), mountain chickadee (Parus gambeli), white-breasted 
nuthatch {Sitta carolinensis), Cassin's finch {Carpodacus 
cassinii), red crossbill {Loxia curvirostra), pine grosbeak 
(Pinicola enucleator), Steller's jay {Cyanocitta stelleri), 
raven (Corvus corax), and red-breasted nuthatch {Sitta 
canadensis). Mammals — chipmimks {Eutamias spp.), 
deer mouse {Peromyscus maniculatus), golden-mantled 
groimd squirrel {Spermophilus lateralis), southern red- 
backed vole {Clethrionomys gapperi), chickaree {Tamia- 
sciurus douglasi) (Hutchins and Lanner 1982; McCaughey 
and Schmidt 1990), and black bears {Ursus americanus) 
(Craighead and others 1982; Kendall 1983). Secondary 
dispersers not only reduce the seed crop but limit the 
availability of seed to nutcrackers and thus the probabil- 
ity of regeneration. 

Usually whitebark pine cone crops are depleted by 
cone and seed consumers. Hutchins and Lanner (1982) 
described a 3-year period when no whitebark pine cones 
fell to the ground other than by animal clipping. In con- 
trast, so many cones were produced in 1989 that many 
were not utilized by seed consumers and fell to the groimd 
(McCaughey and others 1990). Grizzly bears were ob- 
served foraging on these cones in 1990 (Reinhart 1990). 
The 1989 mast year may have reached the upper limit of 



180 



cone production, a level that rarely occiirs. Wind storms 
may have dislodged the whitebark cones making them 
available to ground foraging animals. 

The probability of germination is low from seeds that 
have fallen to the ground in cones or been dropped by ani- 
mals. All whitebark pine seeds sown on the ground are 
eaten or taken by seed consxmiers when not protected 
with exclosures (McCaughey 1990). This situation oc- 
curred in two successive seasons when seeds were sown 
during field germination tests. Eight species of rodents 
were trapped on site dining the field germination test. 
The two most abundant species were the deer mouse and 
the southern red-backed vole representing 54 and 23 per- 
cent of the total popidation trapped (McCaughey 1990). 

GERMINATION 

Information on seed storage and laboratory germina- 
tion of whitebark pine seeds, under a variety of seed treat- 
ment methods, has been summarized and discussed by 
several authors in the symposium proceedings "Whitebark 
Pine Ecosystems: Ecology and Management of a High- 
Mountain Resource" (Schmidt and McDonald 1990). The 
following sections extract pertinent information from 
those various soiirces. 

Seed Storage 

Whitebark pine seed has been frozen imder environ- 
mentally controlled conditions for up to 20 years (Schubert 
1954). Seed viability decreased over that time period from 
50 percent to 3 percent. Mirov (1946) found that viability 
of whitebark pine seed dropped from 24 percent at time 
of collection to 17 percent after being frozen for 8 years, 
but dropped to 1 percent after 11 years of storage. Viabil- 
ity of whitebark pine seed, as related to storage time, is 
dependent on seed maturity when harvested, seed han- 
dling prior to storage, and methods and length of time 
of stratification for germination tests (McCaughey and 
Schmidt 1990). 

Controlled Germination 

ViabiHty of whitebark pine seed, even iinder controlled 
conditions, is highly variable, ranging from 0 to 75 per- 
cent. One procedure for germinating whitebark pine 
seeds is to soak seeds in running tap water for 1 to 2 days 
and stratify moist at 1 to 5 °C for 90 to 120 days in plastic 
bags (Krugman and Jenkinson 1974). Stratification time 
may vary by seed maturity. Jacobs and Weaver (1990a) 
found that 1 month of stratification was sufficient to in- 
crease germination from 5 percent to about 40 percent; 
longer stratification periods (to 8 months) did not improve 
germination. The Coevir d'Alene Nursery in Idaho, 
U.S.A., used the following procedm-es for germination 
tests for western white pine and whitebark pine: 

A. Place seed in nylon mesh bags. 

B. Soak seed for 48 hours in ninning tap water. Place 
nylon mesh bag in plastic bag. 

C. Stratify for 100 days at 1 to 2 °C. Within that 100- 
day stratification time resoak the seed for 1 hour each 
week. 



D. After 100 days, remove seed from stratification and 
surface dry. 

E. Using a vacuum seeder, place 100 seeds on moist 
paper towels (Kimpak) in each of four plastic trays. 

F. Place trays in germinator and take counts. 

Seedling Morphology 

Whitebark pine germinants have thicker stems than 
those of associated conifers. Stem diameters range from 
2 to 4 mm for whitebark and only 1 to 2 mm for lodgepole 
pine {Pinus contorta var. latifolia), Engelmann spruce 
{Picea engelmannii Parry ex Engelm.), and subalpine fir 
iAhies lasiocarpa [Hook.] Nutt.) (McCaughey 1988). Ger- 
minants produce five to 12 cotyledons and grow to heights 
of 3 to 5 cm in the first growing season (fig. 2) (McCaughey 
and Schmidt 1990). Needles vary in length from 7 to 10 
cm with stomata positioned on the dorsal and ventral 
sides. The hypoderm is weak, usually one cell in width; 
the endodermis has an outer wall that is strongly thick- 
ened, and there is only one fibrovascular bundle (Harlow 
1931). 

Whitebark pine trees commonly have multiple stems 
from forking of a single stem or the merging of several 
seedlings from a nutcracker cache. Forking occurs on a 
small percentage of whitebark pine germinants. Nearly 
7 percent of first-year germinants have two forks and 
17 percent have three or more forks (McCaughey 1988; 
Weaver and Jacobs 1990). Genotypic analyses of multi- 
stemmed clumps show that 58, 70, and 83 percent, of 
three stands sampled in Alberta, Canada, and Wyoming, 
U.S.A., had stems of mixed genetic origin while 42, 30, 




Figure 2 — Single whitebark pine seedling germin- 
ating on a mineral seedbed and showing newly 
emerged cotyledons. 



181 



and 17 had arisen by branching (Furnier and others 
1987; Jacobs and Weaver 1990b; Linhart and Tomback 
1985). This indicates that trees of mixed genetic origin 
had merged from several genetically individual seedlings 
from the same cache. 

Whitebark pine germination and root growth are affected 
by early season soil temperatures. Seeds germinate when 
soil temperatures are between 10 and 40 °C with an opti- 
mum temperature range of 25 to 35 °C (Jacobs 1989; 
Jacobs and Weaver 1990a). Root growth occurs between 
10 and 45 °C with the optimum temperature range being 
25 to 35 °C. This optimum range provides conditions 
where whitebark roots grow 5 to 15 mm per day (Jacobs 
and Weaver 1990a). First-year root growth ranges be- 
tween 5 and 18 cm for nutcracker-cached seedlings grow- 
ing in a forest environment (McCaughey 1988). 

Period of Germination 

Germination from a single sowing or caching of white- 
bark pine seeds can continue throughout the growing sea- 
son and during the next 2 to 3 or more years (McCaughey, 
in press b). Because nutcrackers bury seed 2 to 4 cm 
deep, seeds may germinate but not emerge above the soil 
surface for several days after germination or they may not 
emerge at all. The term germinant in this paper refers 
to an emergent, since it is difficult to discern if a seed ger- 
minated yet died prior to emergence. 

Germination Sites 

Whitebark pine germinates on a variety of seedbeds, 
almost entirely dependent on the choice of the nutcracker. 
Whitebark pine grows within 38 habitat-phase combina- 
tions in eastern Idaho and western Wyoming (Steele and 
others 1983) and 44 in Montana (Pfister and others 1977). 
Typical sites for Clark's nutcracker caches are well-drained 
and moist substrates, bare soil, forest litter, gravel, rubble, 
in small plants and logs, in cracks and fissures on exposed 
rock, and in pumice soils (Hutchins and Lanner 1982; 
Tomback 1978, 1982; Tomback and others 1990). Regen- 
eration is commonly found on burned litter seedbeds fol- 
lowing natural and prescribed burns (Morgan and Bimting 
1990; Tomback 1986), 

Nutcrackers cache whitebark seeds on all aspects, but 
the majority of cache sites occur on southeast, south, 
southwest, and west-southwest aspects (Tomback and 
others 1990). Seed storage in south-aspect windblown 
sites probably ensures that some caches are snowfree in 
winter and spring so the nutcracker can retrieve them. 
Even though fewer seeds are cached on the north aspects, 
regeneration densities are highest on north aspects. This 
may be due to less successful retrieval by the nutcracker, 
or it may be a result of favorable environmental condi- 
tions such as moisture and insolation protection during 
the germination period (Tomback and others 1990). The 
tallest and best formed whitebark pine trees are often 
found in high basins or on gentle north slopes (Arno and 



Hoff 1989). Drought and insolation mortality of seedlings 
on south slopes may contribute to the disproportionate 
success of whitebark regeneration on north aspects. 

First-Year Germination 

Whitebark pine seed germinating from caches by the 
nutcracker begins about mid-June and continues through 
early September. The number of whitebark pine seeds 
germinating in the first year after caching varies by year, 
probably in response to timing of spring snowmelt and 
early summer rains (fig. 3) (McCaughey 1990). For ex- 
ample, in 1988 only 11.5 percent of buried seeds germi- 
nated, probably due to below-average precipitation (82 
percent of normal). In 1989 precipitation was near normal 
and germination of 1988 sown seed was nearly 34 percent 
(table 1) (McCaughey 1990). 

First-year germination of whitebark pine is significant- 
ly affected by percent shade cover (0, 25, and 50 percent), 
seedbed condition (mineral, litter, and burned), and sow- 
ing depth (surface sown and buried) (table 1) in controlled 
field studies in southwestern Montana, U.S.A. (McCaughey 
1990). Percent germination of whitebark pine was signifi- 
cantly ip = 0.008) higher (20.4 percent) under a 50 percent 
shade cover than with no shade (16.7 percent) (table 1) 
(McCaughey 1990). Germination of whitebark pine seeds 
sown in 1988 was 26 percent on mineral seedbeds and sig- 
nificantly less on litter and burned seedbeds, 14.5 and 
15.3 percent, respectively (McCaughey 1990). Buried 
whitebark pine seeds (2 to 4 cm deep) had significantly 
higher germination in 1988 and 1989 (11.5 and 33.7 per- 
cent, respectively) than surface-sown seeds that were pro- 
tected from seed consumers in the same years (1.8 and 
11.5 percent, respectively) (McCaughey 1990). 



50- 



? 40- 

a: 

UJ 
(D 

H 30" 
•z. 

UJ 

o 

cc 

lU 
Q. 

lU 
> 



20- 



^ 10- 



3 
O 



Burled, germinating under moist conditions - 1989 



JUNE 




n- 2.700 seeds 



Buried, germinating under dry conditions - 1988 
n- 1,440 seeds 



JULY 



AUGUST ' SEPTEMBER ' OCTOBER 



Figure 3 — Cumulative percent germination of white- 
bark pine germinants from buried and protected 
seeds, sown in 1987 and 1988 respectively, germi- 
nating under dry (1988) and moist (1989) conditions. 
Data collected from a clearcut at 2,530 m elevation 
in southwestern Montana, U.S.A. 



182 



Table 1 — Percent first-year germination of 1987 and 1988 sown 

seeds of whitebark pine as affected by shade cover, seed- 
bed condition, and sowing depth. All treatments excluded 
seed consumers allowing for surface germination. Data 
collected from a clearcut at 2,530 m elevation in south- 
western Montana, U.S.A. 







First-year germination 




Factor 


1988 


1989 


Factor 


level 


mean 


mean 








- Percent 


Shade cover 


0 


^5.2 (a) 


16.7 (c) 


(percent) 


25 


7.9 (a) 


18.8 (cd) 




50 


6.8 (a) 


20.4 (d) 


Seedbed 


Mineral 


8.2 (a) 


26.0 (c) 


condition 


Litter 


5.1 (b) 


14.5 (d) 




Burned 




15.3 (d) 


Sowing depth 


Surface 


1.8 (a) 


3.5 (c) 




Buried 


11.5 (b) 


33.7 (d) 



'Similar and dissimilar letters in parentheses within a column for a factor 
represent nonsignificant and significant differences respectively. 



Delayed Germination 

Delayed germinants are whitebark pine seeds that ger- 
minate two or more seasons after being sown or cached. 
The European stone pine (Pinus cembra) germinates in 
the first, second, and third year after caching (Krugman 
and Jenkinson 1974). Whitebark pine seed germinates 
at least 3 years following sowing (McCaughey 1990; 
McCaughey, in press b; McCaughey and Schmidt 1990). 
Although there were absolute value differences, the effects 
of shade cover and seedbed condition became nonsignifi- 
cant as germination occurred over a 3-year period in a 
study of whitebark germination in western Montana, 
U.S.A. (table 2) (McCaughey, in press b). Germination 
from buried seed (56 percent) remained significantly higher 
than for surface-sown seed (7 percent) 3 years after sow- 
ing (McCavighey, in press b). 

Dormancy of whitebark pine seed is caused by embryo 
underdevelopment, physiological embryo dormancy, im- 
perviousness of the seed coat and female gametophyte 
tissue to oxygen and water uptake, and possibly by depo- 
sition of growth inhibitors to the embryo by female game- 
tophyte tissue (Pitel 1981; Pitel and Wang 1980, 1990). 
Seeds of whitebark pine must mature in a short growing 
season in high-elevation forests, and if climatic conditions 
slow the maturation process nutcrackers will harvest im- 
mature seed. Nutcrackers are normally not concerned 
with seed maturity and will also harvest seeds early in 
the season before they are mature. Embryo vmderdevel- 
opment can be overcome by exposing imbibed seeds to 
20 °C for 30 to 60 days (Leadem 1985). Clipping the seed- 
coat is a method to overcome physiological barriers to ger- 
mination (Pitel 1981). 

Under natural conditions, the greatest germination 
from cached whitebark pine seeds occurs in the second 



year following caching due to delayed dormancy mecha- 
nisms (McCaughey 1990, 1992; Tomback 1992). For 
example, germination of buried whitebark pine seed in- 
creased from 11 percent in the first year following sowing 
to 45 percent in the second year. Germination declined to 
11 percent in the third year foUovnng sowing (McCaughey, 
in press b). 

Limiting Factors 

Germination rates for whitebark pine are highest from 
mid-June through the end of July (McCaughey 1990). 
Mortality of first-year seedlings follows the same pattern; 
the highest mortality rates occvir when germination is 
highest (fig. 4) (McCaughey 1990). Germination exceeds 
mortality up to the first of August, resulting in an accu- 
mulation of surviving seedlings; however, germination 
and mortality rates decrease after the first of August. 
Germination continues from early August until the first 
of September with total nimabers of survivors remaining 
constant because mortality equals germination (McCaughey 
1990). Late germinants have the same probability of svir- 
vival as early germinants. Factors limiting seedling sur- 
vival include: 

Microsite and Biotic — Three major factors are: 
(1) insolation (heat scorching of seedling stem at ground 
surface), (2) drought (drying out of seedling), and (3) ani- 
mals (burial, uprooting, or nipping), specifically pocket 
gophers (Thomomys talpoides) (McCaughey and Schmidt 
1990). Hutchins and Lanner (1982) observed a chipmunk 
uproot and consume a whitebark pine seedling. Insola- 
tion mortality of whitebark pine was higher on mineral, 
litter, and burned seedbeds on nonshaded plots when 
compared to 25 and 50 percent shaded plots in a germina- 
tion test in southwestern Montana, U.S.A. (McCaughey 
1990). 

Blister Rust — The introduced disease, blister rust 
{Cronartium ribicola), poses the most serious threat to 



Table 2 — Percent germination' of 1 987 sown whitebark pine seed 
(buried 2 to 4 cm) as affected by shade cover and seed- 
bed condition for each of the first 3 years following sowing 
and the 3-year total. Data collected from a clearcut at 
2,530 m elevation in southwestern Montana, U.S.A. 



Emergence 



Factor 



Factor 
level 



First 
year 



Second 
year 



Third 
year 



Three-year 
total 



Shade cover 
(percent) 

Seedbed 
condition 



0 

25 
50 

Mineral 
Litter 



10 
13 
12 

13 
9 



Percent 

39 
51 
46 

40 
50 



5 
14 
15 

7 
15 



247 (a) 
61 (a) 
60 (a) 

51 (a) 
60 (a) 



'Percent germination for each year is based on the number of seeds that 
had not previously emerged. 

^Similar and dissimilar letters in parentheses for a factor represent statisti- 
cally nonsignificant and significant differences respectively. 



183 



10 
8 

6 -I 



1988 



INSOLATION 
DROUGHT 



I 



JUNE 



JULY AUGUST SEPTEMBER 



1989 




■1 INSOLATION 




m DROUGHT 






□ ANIMAL 








JUNE 1 


JULY 1 AUGUST 1 SEPTEMBER 



Figure 4 — Percent first-year mortality of whitebark 
pine seedlings by causal agent over time for 1 987 
and 1 988 sown seeds. Data collected from a clear- 
cut at 2,530 m elevation in southwestern Montana, 
U.S.A. 



The Yellowstone environment is poorly suited to blister 
rust infection because it has a cool, dry climate that is 
only marginally suitable for blister rust teliospore germi- 
nation (Krebill 1971). 

Mountain Pine Beetle — Mountain pine beetle is the 
most damaging insect in mature stands of whitebark pine 
(Arno and Hoff 1989; McCaughey and Schmidt 1990). 
Large numbers of whitebark pine are killed by moimtain 
pine beetle when epidemics spread into whitebark pine 
stands from the lower elevation lodgepole pine zone. These 
epidemics, like blister rust, create conditions where suc- 
cession progresses toward dominance by shade-tolerant 
conifers such as subalpine fir and Engelmann spruce (Arno 
1986). 

Fire Suppression — Whitebark pine is serai to subal- 
pine fir and other conifers, and periodic fires have helped 
to perpetuate the pine (Arno 1986). Past fire intervals 
in whitebark pine stands ranged from 50 to 300 years or 
more depending on site conditions (Arno 1980). Fires oc- 
cur in whitebark pine stands only under severe burning 
conditions and are tj^ically non-stand replacing. Fires 
sweep across large areas of forest reducing competition 
from shade-tolerant conifers (Arno 1976). Nearly 90 years 
of fire suppression have reduced the annual acreage of 
whitebark pine forests being biu'ned, creating conditions 
where shade-tolerant species have taken over what were 
once serai whitebark pine stands. 

Severe fire conditions are being created by increased 
mortality and the buildup of fuels in whitebark pine stands 
due to mountain pine beetle and blister rust epidemics. 
Stand replacement fires will be the result of this unnatu- 
ral fuel loading and will eliminate large cone-producing 
stands of whitebark pine. 



the survival of whitebark pine in parts of its distribu- 
tional range. Blister rust was brought to western North 
America in 1910 on a boatload of blister rust-infected 
eastern white pine (Pinus strobus L.). Whitebark pine is 
the most susceptible to blister rust infection of 14 white 
pines rated by Bingham (1972). Blister rust causes exten- 
sive damage and mortality in whitebark pine stands in 
moist mountain regions in northwestern Montana, north- 
ern Idaho, and the Washington Cascades in the United 
States (Arno 1986). Damage is light where environmental 
conditions are somewhat dry and cold, and there is a de- 
crease in low-elevation sources of inoculimi (Hoff and 
Hagle 1990). 

Blister rust-infected trees may live for several years 
before mortality occurs; during that time cone production 
is reduced due to loss of upper crowns. Mortality from 
rust favors succession toward dominance by whitebark 
pine's shade-tolerant associates thus reducing site poten- 
tial for whitebark pine regeneration and survival (Arno 
1986; Arno and Hoff 1989; McCaughey and Schmidt 1990). 

The degree of blister rust infection on whitebark pine 
decreases southward for all parts of its range (Hoff and 
Hagle 1990; Keane and others 1990). A major population 
of whitebark pine exists in the Yellowstone ecosystem of 
southwestern Montana and northwestern Wyoming, U.S.A. 



VEGETATIVE REPRODUCTION 

Whitebark pine can reproduce vegetatively through lay- 
ering of lower branches along the ground surface. In the 
Mission Range of western Montana vegetative reproduc- 
tion was observed from layering of shrublike (krummholz) 
whitebark pine (Arno 1981; Arno and Hoff 1989). Although 
layering is possible, the vast majority of reproduction is 
from seeds. 

SUMMARY 

Regeneration processes are similar for whitebark pine 
and the other stone pines of the world. The initiation, 
development, and maturation of cones and seeds for white- 
bark pine are characteristic of five-needle pines in general 
with seasonal timing of these processes varying between 
species. The entire process from cone initiation to cone 
and seed maturity takes about 24 months. Cone and seed 
development are most affected by variations in climatic 
conditions during periods of reproductive bud formation, 
pollination, and growth and development. Several insects 
and diseases reduce survival of cones and seeds of white- 
bark pine. 

Regeneration of whitebark pine depends almost exclu- 
sively on the seed selection, dispersal, and caching habits 



184 



of the Clark's nutcracker. Other stone pines of the world 
have similar mutualistic relationships with bird species 
for seed dispersal. 

Whitebark pine seeds are cached by Clark's nutcrackers 
on a variety of sites ranging from forest litter to cracks 
and fissures in rocks. The majority of regeneration occurs 
on south or west aspects. Seed storage in south-aspect 
windblown sites ensures that some caches are snow free 
in winter and spring for nutcracker retrieval. Regenera- 
tion densities are highest on north aspects even though 
fewer seeds are cached there because of favorable micro- 
site conditions such as moistxu'e and insolation protection. 

Germination of whitebark pine begins about mid-June 
and continues through early September in the first year 
following caching. Delayed germination occurs for up 
to 3 or 4 years after caching. Total germination over a 
3-year period is highest on litter seedbeds that have 50 
percent shade cover, while germination is lower on min- 
eral and nonshaded seedbeds (McCaughey, in press b). 
Dormancy of whitebark pine seed is caused by embryo 
underdevelopment, physiological embryo dormancy, and 
imperviousness of the seed coat and female gametoph3^e 
tissue to oxygen and water uptake. 

Factors limiting survival of whitebark pine seedlings 
are microsite, biotic, disease, insects, and fire. Insolation 
and drought are microsite-associated factors that cause 
seedling mortality. Animals such as pocket gophers bury, 
uproot, or nip off young seedlings, and chipmunks uproot 
and consimie whole seedlings. Deer mice and southern 
red-backed voles and other small mammals consume seed 
before they germinate. Blister rust causes mortality or 
extensive damage of the cone-bearing portions of the tree. 
Whitebark pine is killed by the mountain pine beetle 
when epidemics spread into whitebark pine stands from 
the lower elevation lodgepole pine zone. Fire suppression 
creates conditions where shade-tolerant species such as 
Engelmann spruce and subalpine fir, normally killed by 
light imderburns, take over through succession what were 
once serai whitebark pine stands. Severe fire conditions 
are being created by the buildup of fuels in whitebark 
pine stands due to increased mortality. 

We are beginning to piece together the puzzle of the re- 
generation process of whitebark pine. Survival of wildlife 
species such as the grizzly is threatened due to a reduc- 
tion or loss of localized populations of whitebark pine 
caused by global climate change, introduced disease, in- 
sects, and succession fi-om fire exclusion. Future research 
should define microsite and biotic factors influencing ger- 
mination and long-term survival of seedlings. Informa- 
tion on regeneration processes from other stone pine spe- 
cies from around the world is helping identify key areas 
for research on whitebark pine. For example, little infor- 
mation is available on the elevational distribution of 
whitebark and which habitats are best suited for regen- 
eration. Cembra pine {Pinus cembra) grows in the Swiss 
Alps fi-om valley bottoms to timberline when management 
actions reduce intertree competition. Whitebark pine 
management wiU improve from increased knowledge of 
regeneration processes of all the stone pines. 



REFERENCES 

Amo, S. F. 1976. The historical role of fire on the Bitter- 
root National Forest. Res. Pap. INT-187. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Forest and Range Experiment Station. 30 p. 

Arno, S. F. 1980. Forest fire history in the Northern 
Rockies. Journal of Forestry. 78(8): 460-465. 

Arno, S. F. 1981. [Unpublished observations]. Missoula, 
MT: U.S. Department of Agriculture, Forest Service, 
Intermoimtain Fire Sciences Laboratory. 

Amo, S. F. 1986. Whitebark pine cone crops - a diminish- 
ing source of wildlife food? Western Journal of Applied 
Forestry. 1(3): 92-94. 

Arno, S. F.; Hoff, R. J. 1989. Silvics of whitebark pine 
(Pinus albicaulis). Gen. Tech. Rep. INT-253. Ogden, 
UT: U.S. Department of Agriculture, Forest Service, 
Intermoimtain Research Station, lip. 

Bartos, D. L.; Gibson, K. E. 1990. Insects of whitebark 
pine with emphasis on mountain pine beetle. In: 
Schmidt, Wyman C; McDonald, Kathy J., comps. 
Proceedings — symposium on whitebark pine ecosys- 
tems: ecology and management of a high-mountain re- 
source; 1989 March 29-31; Bozeman, MT. Gen. Tech. 
Rep. INT-270. Ogden, UT: U.S. Department of Agricul- 
ture, Forest Service, Intermountain Research Station: 
171-178. 

Bingham, R. T. 1972. Taxonomy, crossability, and relative 
blister rust resistance of 5-needled white pines. In: Biol- 
ogy of rust resistance in forest trees: proceedings of a 
NATO-IUFRO advance study institute; 1969 August 
17-24; Moscow, ID. Misc. Publ. 1221. Washington, DC: 
U.S. Department of Agriculture: 271-280. 

Bock, W. J.; Balda, R. P.; Vander Wall, S. B. 1973. Mor- 
phology of the sublingual pouch and tongue muscula- 
ture in Clark's nutcracker. Auk. 90: 491-519. 

Craighead, J. J.; Sumner, J. S.; Scaggs, G. B. 1982. A de- 
finitive system for analysis of grizzly bear habitat and 
other wilderness resources. Wildlands Inst. Monogr. 1. 
Missoula, MT: University of Montana. 279 p. 

Dewey, Jerald E. 1989. [Personal communication to Dale 
Bartos of Intermountain Research Station, Logan, UT]. 
March 27. Missoula, MT: U.S. Department of Agricul- 
ture, Forest Service, Northern Region. 

Edwards, D. G. W. 1990. Cone prediction, collection and 
processing. In: Schmidt, Wyman C; McDonald, Kathy J., 
comps. Proceedings — symposium on whitebark pine eco- 
systems: ecology and management of a high-mountain 
resource; 1989 March 29-31; Bozeman, MT. Gen. Tech. 
Rep. INT-270. Ogden, UT: U.S. Department of Agricul- 
ture, Forest Service, Intermountain Research Station: 
78-102. 

Eggers, D. E. 1986. Management of whitebark pine as po- 
tential grizzly bear habitat. In: Contreras, G. P.; Evans, 
K. E., comps. Proceedings — grizzly bear habitat sympo- 
sium; 1985 April 30-May 2; Missoula, MT. Gen. Tech. 
Rep. INT-207. Ogden, UT: U.S. Department of Agricul- 
ture, Forest Service, Intermountain Research Station: 
170-175. 



185 



Furnier, G.; Knowles, P.; Clyde, M.; Dancik, B. 1987. Ef- 
fects of avian seed dispersal on the genetic structure 
of whitebark pine populations. Evolution. 41: 607-612. 

Harlow, W. M. 1931. The identification of the pines of the 
United States, native and introduced, by needle struc- 
ture. Tech. Publ. 32. Syracuse, NY: New York State 
College of Forestry. 21 p. 

Hoff, R. J. 1992. Unpublished data on file at: U.S. Depart- 
ment of Agriculture, Forest Service, Intermountain Re- 
search Station, Forestry Sciences Laboratory, Moscow, 
ID. 

Hoff, R. J.; Hagle, S. 1990. Diseases of whitebark pine 
with special emphasis on white pine blister rust. In: 
Schmidt, Wyman C; McDonald, Kathy J., comps. 
Proceedings — symposium on whitebark pine ecosys- 
tems: ecology and management of a high-mountain 
resource; 1989 March 29-31; Bozeman, MT. Gen. Tech. 
Rep. INT-270. Ogden, UT: U.S. Department of Agricul- 
ture, Forest Service, Intermountain Research Station: 
179-190. 

Hutchins, H. E.; Lanner, R. M. 1982. The central role of 
Clark's nutcracker in the dispersal and establishment 
of whitebark pine. Oecologia. 55: 192-201. 

Jacobs, J. S. 1989. Temperature and light effects on seed- 
ling performance of Pinus albicaulis. Bozeman, MT: 
Montana State University. 39 p. Thesis. 

Jacobs, J.; Weaver, T. 1990a. Effects of temperature and 
temperature preconditioning on seedling performance 
of whitebark pine. In: Schmidt, Wyman C; McDonald, 
Kathy J., comps. Proceedings — symposium on white- 
bark pine ecosystems: ecology and management of a 
high-mountain resource; 1989 March 29-31; Bozeman, 
MT. Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Depart- 
ment of Agriculture, Forest Service, Intermountain 
Research Station: 134-139. 

Jacobs, J.; Weaver, T. 1990b. Occurrence of multiple 
stems in whitebark pine. In: Schmidt, Wyman C; 
McDonald, Kathy J., comps. Proceedings — symposium 
on whitebark pine ecosystems: ecology and manage- 
ment of a high-moimtain resource; 1989 March 29-31; 
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Research Station: 156-159. 

Keane, R. E.; Arno, S. F.; Brown, J. K.; Tomback, D. F. 
1990. Simulating disturbances and conifer succession 
in whitebark pine forests. In: Schmidt, Wyman C; 
McDonald, Kathy J., comps. Proceedings — symposium 
on whitebark pine ecosystems: ecology and manage- 
ment of a high-mountain resource; 1989 March 29-31; 
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Research Station: 274-288. 

Kendall, K. C. 1983. Use of pine nuts by grizzly and black 
bears in the Yellowstone area. In: International Confer- 
ence on Bear Research and Management. Madison, WI: 
International Bears Association. 5: 166-173. 

Knight, R. R.; Blanchard, B. M.; Mattson, D. J. 1987. 
Yellowstone grizzly bear investigation: report of the 
interagency study team, 1987. Bozeman, MT: U.S. De- 
partment of the Interior, National Park Service. 80 p. 



Krebill, R. G. 1971. Conditions for teliospore germination 
in Yellowstone National Park environment. Unpub- 
lished report on file at: U.S. Department of Agriculture, 
Forest Service, Intermountain Research Station, 
Ogden, UT. 

Krugman, S. L.; Jenkinson, J. L. 1974. Pinus L. pine. 
In: Schopmeyer, C. S., ed. Seeds of woody plants in the 
United States. Agric. Handb. 450. Washington, DC: 
U.S. Department of Agriculture, Forest Service: 598-638. 

Landis, T. D.; Tinus, R. W.; McDonald, S. E.; Barnett, 
J. P. 1990. The biological component: nursery pests and 
mycorrhizae. Vol. 5. The container tree nursery manual. 
Agric. Handb. 674. Washington, DC: U.S. Department 
of Agriculture, Forest Service. 171 p. 

Lanner, R. M. 1980. Avian seed dispersal as a factor in 
the ecology and evolution of limber and whitebark 
pines. In: Sixth North American forest biology work- 
shop proceedings; 1980 August; Edmonton, AB. 
Edmonton, AB: University of Alberta: 14-48. 

Lanner, R. M. 1982. Adaptations of whitebark pine for 
seed dispersal by Clark's nutcracker. Canadian Journal 
of Forest Research. 12(2): 391-402. 

Lanner, R. M.; Vander Wall, S. B. 1980. Dispersal of lim- 
ber pine seed by Clark's nutcracker. Journal of For- 
estry. 78(10): 637-639. 

Leadem, C. L. 1985. Seed dormancy in three Pinus species 
of the Inland Mountain West. In: Schmidt, Wyman C, 
comp. Proceedings — future forests of the Moimtain 
West: a stand culture symposium; 1986 September 29- 
October 3; Missoula, MT. Gen. Tech. Rep. INT-243. 
Ogden, UT: U.S. Department of Agriculture, Forest 
Service, Intermountain Research Station: 117-124. 

Linhart, Y.; Tomback, D. F. 1985. Seed dispersal by nut- 
crackers causes multi-trunk growth form in pines. 
Oecologia. 67: 107-110. 

McCaughey, W. W. 1990. Biotic and microsite factors 
affecting Pinus albicaulis establishment and survival. 
Bozeman, MT: Montana State University. 78 p. 
Dissertation. 

McCaughey, W. W. 1988. Unpublished data on file at: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Research Station, Forestry Sciences Labora- 
tory, Bozeman, MT. 

McCaughey, W. W. 1992. Unpublished data on file at: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Research Station, Forestry Sciences Labora- 
tory, Bozeman, MT. 

McCaughey, W. W. [In press a]. Seasonal maturation of 
whitebark pine seed in the Greater Yellowstone ecosys- 
tem. In: Schullery, Paul; Despain, Don, comps. Proceed- 
ings of the 1st biennial scientific conference on the 
Greater Yellowstone Ecosystem — plants and their envi- 
ronments; 1991 September 16-17; Mammoth Hot 
Springs, Yellowstone National Park, WY. Tech. Rep. 
USDI National Park Service. 

McCaughey, W. W. [In press b]. Delayed germination and 
seedling emergence of Pinus albicaulis in a high eleva- 
tion clearcut in Montana, U.S.A. In: Edwards, D. G., 
comp. Proceedings — symposium: seed dormancy and 
barriers to germination; 1991 April 23-26; lUFRO 



186 



project group P2. 04-00 meeting on seed problems; 
Victoria, BC. Victoria, BC: Forestry Canada. 

McCaughey, W. W.; Amo, S. F.; Tomback, D. F. 1990. 
Unpublished personal observations from Henderson 
Mountain near Cooke City, MT, U.S.A.. On file at: U.S. 
Department of Agriciilture, Forest Service, Intermoun- 
tain Research Station, Forestry Sciences Laboratory, 
Bozeman, MT. 

McCaughey, W. W.; Schmidt, W. C. 1990. Autecology of 
whitebark pine. In: Schmidt, Wjonan C; McDonald, 
Kathy J., comps. Proceedings — symposium on whitebark 
pine ecosystems: ecology and management of a high- 
mountain resource; 1989 March 29-31; Bozeman, MT. 
Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Department 
of Agriculture, Forest Service, Intermoimtain Research 
Station: 85-96. 

Mirov, N. T. 1946. Viability of pine seed after prolonged 
cold storage. Journal of Forestry. 44(3): 193-195. 

Mirov, N. T. 1967. The genus Pinus. New York: The Ronald 
Press. 602 p. 

Morgan, Penny; Bunting, Stephen C. 1990. Fire effects 
in whitebark pine forests. In: Schmidt, Wyman C; 
McDonald, Kathy J., comps. Proceedings — symposiimi 
on whitebark pine ecosystems: ecology and manage- 
ment of a high-mountain resource; 1989 March 29-31; 
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Research Station: 166-170. 

Pfister, R. D.; Kovalchik, B. L.; Arno, S. F.; Presby, R. C. 
1977. Forest habitat types of Montana. Gen. Tech. Rep. 
INT-34. Ogden, UT: U.S. Department of Agriculture, 
Forest Service, Intermountain Forest and Range Ex- 
periment Station. 174 p. 

Pitel, J. A. 1981. Improved germination of whitebark 
pine {Pinus alhicaulis Engelm.) seeds. Chalk River, ON: 
Canadian Forestry Service, Petawawa National For- 
estry Institute; unpublished report. 5 p. 

Pitel, J. A.; Wang, B. S. P. 1980. A prehminary study of 
dormancy in Pinus albicaulis seeds. Bi-Monthly Re- 
search Notes. 36(1): 4-5. 

Pitel, J. A.; Wang, B. S. P. 1990. Physical and chemical 
treatments to improve germination of whitebark pine 
seeds. In: Schmidt, Wyman C; McDonald, Kathy J., 
comps. Proceedings — symposium on whitebark pine eco- 
systems: ecology and management of a high-movmtain 
resource; 1989 March 29-31; Bozeman, MT. Gen. Tech. 
Rep. INT-270. Ogden, UT: U.S. Department of Agriciil- 
ture, Forest Service, Intermoimtain Research Station: 
130-133. 

Reinhart, D. P. 1990. [Personal commmiication]. August. 
Bozeman, MT: U.S. Department of the Interior; 
Montana State University, Interagency Grizzly Bear 
Project. 

Reinhart, D. P.; Mattson, D. J. 1990. Red sqiiirrels in the 
whitebark zone. In: Schmidt, Wyman C; McDonald, 
Kathy J., comps. Proceedings — symposium on whitebark 
pine ecosystems: ecology and management of a high- 
moxmtain resource; 1989 March 29-31; Bozeman, MT. 
Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Department 
of Agriculture, Forest Service, Intermoimtain Research 
Station: 256-263. 



Schmidt, W. C; Lotan, J. E. 1980. Phenology of common 
forest flora of the Northern Rockies— 1928 to 1937. Res. 
Pap. INT-259. Ogden, UT: U.S. Department of Agricul- 
ture, Forest Service, Intermountain Forest and Range 
Experiment Station. 20 p. 

Schmidt, W. C; McDonald, K. J., comps. 1990. 
Proceedings — symposium on whitebark pine eco- 
systems: ecology and management of a high-mountain 
resource; 1989 March 29-31; Bozeman, MT. Gen. Tech. 
Rep. INT-270. Ogden, UT: U.S. Department of Agricul- 
ture, Forest Service, Intermountain Research Station. 
386 p. 

Schubert, G. H. 1954. Viability of various coniferous seeds 
after cold storage. Journal of Forestry. 52: 446-447. 

Smith, C. C. 1968. The adaptive nature of social organiza- 
tion in the genus of tree squirrels Tamiasciurus . Eco- 
logical Monographs. 38: 31-63. 

Steele, R.; Cooper, S. V.; Ondov, D. M.; Roberts, D. W.; 
Pfister, R. D. 1983. Forest habitat types of eastern 
Idaho-western Wyoming. Gen. Tech. Rep. INT-144. 
Ogden, UT: U.S. Department of Agriculture, Forest 
Service, Intermountain Forest and Range Experiment 
Station. 122 p. 

Tomback, D. F. 1978. Foraging strategies of Clark's nut- 
cracker. Living Bird. 16(1977): 123-160. 

Tomback, D. F. 1982. Dispersal of whitebark pine seeds 
by Clark's nutcracker: a mutualism hypothesis. Journal 
of Animal Ecology. 51(2): 451-467. 

Tomback, D. F. 1983. Nutcrackers and pines: coevolution 
or coadaptation? In: Nitecki, M. H., ed. Coevolution. 
Chicago and London: University of Chicago Press: 
179-223. 

Tomback, D. F. 1986. Post-fire regeneration of krummholz 
whitebark pine: a consequence of nutcracker seed cach- 
ing. Madrono. 33(2): 100-110. 

Tomback, D. F. 1992. Unpublished data on file at: Depart- 
ment of Biology, University of Colorado at Denver, 
Denver, CO. 

Tomback, D. F.; Hoffmann, L. A.; Sund, S. K. 1990. Co- 
evolution of whitebark pine and nutcrackers: implica- 
tions for forest regeneration. In: Schmidt, Wyman C; 
McDonald, Kathy J., comps. Proceedings — symposiima 
on whitebark pine ecosystems: ecology and manage- 
ment of a high-mountain resource; 1989 March 29-31; 
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Research Station: 118-129. 

Vander Wall, S. B.; Balda, R. P. 1977. Coadaptations of 
the Clark's nutcracker and the pinon pine for efficient 
seed harvest and dispersal. Ecological Monographs. 
47: 89-111. 

Weaver, T.; Jacobs, J. 1990. Occurrence of multiple stems 
in whitebark pine. In: Schmidt, Wjonan C; McDonald, 
Kathy J., comps. Proceedings — symposium on white- 
bark pine ecosystems: ecology and management of a 
high-mountain resource; 1989 March 29-31; Bozeman, 
MT. Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Depart- 
ment of Agriculture, Forest Service, Intermountain Re- 
search Station: 156-159. 



187 



JAPANESE STONE PINE CONE 
PRODUCTION ESTIMATED FROM CONE 
SCARS, MOUNT KISOKOMAGATAKE, 
CENTRAL JAPANESE ALPS 

Ikuko Nakashinden 



Abstract — The past cone production of the Japanese stone pine 
(Pinus pumila) on Mount Kisokomagatake was reconstructed 
from cone scars. The main cone production was by larger indi- 
viduals in the lower part of the scrub zone; it was greater than 
that in the upper part where growth and crop are controlled 
mainly by strong winter winds and snow. 



The Japanese stone pine {Pinus pumila Regal) is a 
genetic dwarf pine distributed over the eastern part of 
Siberia and China, the northern part of Korea, and Japan. 
In Japan, most of its habitats are located on high moun- 
tains. The climax scrub zone is formed above the timber- 
line of subalpine coniferous forests. At the southern mar- 
gin of its range (Chubu District) the scrub zone occupies 
the altitudinal zone from 1,961 to 3,192 m (Yanagimachi 
and Ohmori 1991). 

The Japanese stone pine propagates in two different 
ways: one by the expansion of shoots and the other by 
seeds. The latter is the most important for spatial ex- 
pansion (Okitsu 1990). Seeds are dispersed by seed- 
storing corvids {Nucifraga caryocatactes) (Hayashida 
1989; Tomback 1990). Studies concerning pine cone 
production have begun recently. Okitsu and Mizoguchi 
(1990, 1991) evaluated pine cone productivity based on 
the individual growth sizes and scrub sizes. In addition, 
studies of the relationship between environment of 
the habitats and cone production have commenced 
(Nakashinden 1990). These studies are limited to 1 year's 
cone production data. It is notable that past cone produc- 
tion can be estimated from cone scars. This was done on 
whitebark pine {Pinus albicaulis) in the United States 
(Morgan and Bimting 1992; Weaver and Forcella 1986). 
Saito and others (1989) reported that cone scars are also 
recognizable in the Japanese stone pine in Hokkaido, but 
little information about it was given in their report. 

This author developed the method to reveal the yearly 
change in cone production for the Japanese stone pine and 
succeeded in reconstructing pine cone production over the 
past 15 years (Nakashinden 1991). Using this cone scar 
method, chronological pine cone production of the Japa- 
nese stone pine is described and the relationship to the 
environment of its habitats is examined. 



STUDY AREA AND METHODS 

The study area is located on Moimt Kisokomagatake, 
whose simimit is 2,956 m, in the Central Japanese Alps, 
in Chubu District (fig. 1). This location is one of the 
southernmost parts of the Japeinese stone pine 
distribution. 

The habitats examined are located on the mountain 
slopes between 2,600 and 2,950 m elevation, including 
the summits of Mount Kisokomagatake and Mount 
Chausudake (2,652 m). In this altitudinal zone, the 
Japanese stone pine is dominant, forming its own vege- 
tation zone above the subalpine forest of Abies veitchii, 
Tsuga diversifolia, Betula ermanii, and Alnus 
maximowiczii. 




Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Ikuko Nakashinden is Graduate Student, Department of Geography, 
University of Tokyo, Bunkyo-ku, Tokyo 113, Japan. 



Figure 1— Location of Mount Kisokomagatake. 



188 



Pine Cone Scars 

The positions of pine cones of the Japanese stone pine 
are restricted to the first nodes of the annual shoots. 
They matiire in the second year after fruiting. In 
September, nutcrackers harvest cones of the Japanese 
stone pine as a food source. When the pine cone is picked, 
a scar of the cone remains on the stem (Nakashinden 
1991; Saito and others 1989). Each scar represents one 
cone. From these scars, the number of pine cones can be 
calciilated. The production can be determined from the 
location of the nodes on the stems. This cone scar method 
was used on whitebark pine in the United States, and the 
scars could be measured for 6 to 8 years (Weaver and 
Forcella 1986), or 6 to 12 years (Morgan and Bunting 
1992). For the Japanese stone pine, the scars could be 
counted for 15 to 20 years (Nakashinden 1991). 

Using this method, 32 plots (1 m^) at different altitudes 
from 2,600 m near the forest hne to 2,950 m near the 
summit were investigated. Ten stems in each plot were 
selected for counting pine cone scars and for the measure- 
ment of individual stem growth for the 15 years preceding 
1990. 

Distribution, Geomorphology, and 
Snow 

Using color aerial photographs and field observation, 
the Japanese stone pine scrubs in the study area were 
delineated and mapped. Some topographical maps were 
made during the field survey in order to examine the mi- 
croscale landform beneath the scrubs. The snow melting 
status and snow melting period near the simimit of 
Moxmt Kisokomagatake was observed and recorded on 
photographs in each spring season from 1989 to 1991. 



250 - 

200 - 



0) 

o 

0 



150 - 



100 - 



50- 



r— T 

76 77 78 79 80 81 82 

Year 



II I ' I I I I 

83 84 85 86 87 88 89 90 



Figure 2 — Annual cone crops of Japanese stone 
pine (1976-90). Cone crops were reconstructed 
from cone scars on the stems. Total number from 
32 plots is shown for each year. 



V) 

o 
o 
E 

C 

o 
O 



4.0 



3.0 



2.0 



1.0 




1983 



J L 



J L 



76 



78 



80 



82 84 

Year 



86 88 



90 



Figure 3 — Mean annual stem elongation of 
Japanese stone pine (1976-90). Ten stems 
each in 32 plots were measured. 



CONE PRODUCTION OVER 15 YEARS 

The past 15 years of pine cone production was recon- 
structed on the study area (fig. 2). Mast years were 
recognized in 1981, 1984, 1988, and 1989, showing a 
3- or 4-year periodicity of cone production. The lowest 
production during the past 15 years occurred in 1982 and 
1983. In 1983, the annual shoot length was also the low- 
est during the 15-year period (fig. 3). The year 1983 is 
judged to be one of the worst years for both cone crop and 
growth. 

On the individual plots, both mast and fail years oc- 
curred regardless of their environmental differences. 
For the mast years, 68 percent of the 32 plots showed 
high production in 1981, and in 1984 it was 65 percent. 
For the fail years, 59 percent of the plots failed in cone 
production in 1982, and in 1983 it was 63 percent. These 
occurrences were dependent on the climate and not on 
the growth forms or the location of scrubs. 

DISTRffiUTION PATTERNS BY 
ALTITUDE 

On Mount Kisokomagatake, the Japanese stone pine 
scrubs dominate and form a vertical zone above the subal- 
pine forest at an elevation between 2,600 and 2,950 m. 
However, the distribution patterns were characterized by 
the altitude. On the upper part from 2,850 up to 2,950 m, 
near the summit of Moimt Kisokomagatake, the distribu- 
tion pattern of the scrubs shows a stripe pattern (fig. 4). 

The scrubs are fragmented by huge granite rocks, tors, 
stone bank terraces, trails, and alpine plant communities 
such as Arctous alpinus vai.Japonicus, Diapensia lap- 
ponia var. obovata, Oxytropis japonica Maxim, Leonto- 
podium shinanense Kitam, Gentiana algida Pall., and 
Loiseleuria procumhens Desv. The lower part between 
2,850 and 2,600 m, near the boimdary to the subalpine 
forest shows an extensively spreading scrub, covering the 
slopes entirely from the ridges down to the forest limit 
(fig. 5). 



189 




Figure 4 — Distribution pattern of Japanese stone 
pine scrubs in tine upper part of the Mount 
Kisokomagatake study area. Distribution is frag- 
mented by huge rocks, tors, stone bank terraces, 
trails, and alpine plant communities. 



GROWTH FORM AND CONE 
PRODUCTION RELATED TO 
ALTITUDE 

Growth form and pine cone production are ditTerent be- 
tween the upper and the lower part of the Japanese stone 
pine scrub zone (table 1). The mean scrub height, stem 
length, and diameter at the stem base in the lower part 
are almost twice as large as those in the upper part. The 
mean pine cone production per stem is 1.5 times more 
than in the upper part. Although the niunber of stems 
per square meter in the upper part was twice that of the 
lower part, the Japanese stone pine scrubs in the lower 
part were larger in size and produced more cones than the 
scrubs in the upper part. 



DISCUSSION AND CONCLUSIONS 

On Moimt Kisokomagatake, the mean summer tem- 
perature (July and August) was 10 to 12 °C at 2,600 m 
(Koizumi 1974) and 10 °C (Nakashinden 1990, unpub- 
lished) at 2,850 m. Concerning the lower parts of the 
Japanese stone pine scrub zone, the thermal conditions 
indicate that the Japanese stone pine scrub zone is in- 
cluded in the thermal subalpine zone (Yanagimachi and 
Ohomori 1991). There, the Japanese stone pine scrubs 
are considered to grow and produce cones at the same 
rates as in the subalpine zone. 

The scrubs in the upper part are explained by other en- 
vironmental factors unfavorable for growth and cone pro- 
duction. In the Japanese stone pine scrub zone, the scrub 
height is related to winter wind strength and snow depth 
of their habitats (Okitsu 1987). In high-moimtain areas, 
the Japanese stone pine needs to be protected by snow- 
pack to evade physical damages or dehydration caused by 
strong winds. Short scrub height indicates strong winds 
and shallow snow cover at the habitats during the winter 
season, while tall scrub height indicates moderate winds 
and heavy snowpack. 

On Movmt Kisokomagatake, the depth of snow varies 
from 10 cm to 3 m according to landforms and wind veloc- 
ity (Koiziuni 1974). Especially around the simimit, most 
of the snow is blown away from the windward slope by 
strong westerly winds (fig. 6). Snow accumulates on the 
leeward side of the ridges or rocks. On the windward 
slope, characterized by shallow snow cover, the Japanese 
stone pine grow close to the ground, thus protecting them- 
selves from strong winds (fig. 7). Therefore, vertical 
growth of the Japanese stone pine scrubs is disturbed by 
winter winds. 

Variation in snow depth caused differences in the 
length of the snow melting period. According to 3 years' 
observation by this author, snow on the Japanese stone 
pine scrubs located on windward slopes started to melt 




Figure 5 — Distribution pattern of Japanese 
stone pine scrubs in the lower part of the 
Mount Kisokomagatake study area. They 
spread extensively and cover the slopes en- 
tirely from the ridges down to the forest line. 



Table 1 — The growth and pine cone production in each part of 
the distribution zone viewed from different altitude of 
quadrats. The boundary of the upper and the lower part 
is 2,850 m 



Cone production and growth 


Upper part 


Lower part 


Cone production (pieces) 


4.1 


6.4 


Mean scrub height (cm) 


36.2 


89.2 


Mean stem length (cm) 


92.0 


180.0 


Mean diameter at stem base (cm) 


2.2 


4.2 


Mean annual stem elongation (cm) 


2.4 


3.0 


Number of stems/m^ 


38.5 


20.0 



'The values indicate the means of 20 plots for the upper part and 12 plots 
for the lower part. 



190 




Figure 6 — Mount Kisokomagatake summit in 
early spring (1 991 ). Snow depth varies accord- 
ing to wind velocities and topography. The 
western-most part (left) where snow cover was 
thin due to strong westerly winds shows earliest 
snow melt in spring. 



at the end of April. On the leev^^ard slopes, the scrubs 
became completely free of snow by the middle of June. 
On the slopes where snow remains imtil the end of July, 
the Japanese stone pine is replaced by alpine snowbed 
vegetation. On the other hand, shallow snow cover causes 
early snow melt and solifluction is very common. Solifluc- 
tion forms stone banked terraces and cuts off the roots of 
young plants. For this reason, the Japanese stone pine 
scrubs in the upper part are not able to cover all the 
slopes and thus show a fragmented distribution pattern. 

These severe environmental conditions are not favor- 
able for growth or cone production of the Japanese stone 
pine. The main cone production is performed by large and 




Figure 7 — Creeping Japanese stone pine in the 
upper part of the study area. The scrubs' height is 
10-15 cm. The stem has branches flagged toward 
the leeward side. The plot shows one of the low- 
est cone productions (only one cone for 1 0 stems) 
during these 1 5 years. 



tall scrubs at the lower elevations where winter winds are 
moderate and snow is evenly distributed. 

ACKNOWLEDGMENTS 

I thank Professor Dr. F.-K. Holtmeier, Dr. Wyman C. 
Schmidt, and Dr. W. Schonenberger for their invitation 
to attend the workshop. I thank Dr. Professor Hiroo 
Ohmori and Dr. Takeei Koizumi for their valuable advice 
in carrying out this study. My thanks also go to Dr. Tad 
Weaver and Dr. Penelope Morgan for providing me with 
information about cone scar methods on whitebark pine. 
Dr. Mitsuhiro Hayashida offered me valuable information 
in preparation for the workshop. I am grateful to all my 
colleagues in the Geography Department of Tokyo 
Gakugei University, who enthusiastically helped me in 
my field research on Mount Kisokomagatake. I thank 
Mr. Wataru, Mrs. Joyce Mukai, and Mr. Garrett Vance 
who checked my manuscript. And I thank Dr. Miki 
Kobayashi who always encourages me to keep studying 
Pinus pumila. 

REFERENCES 

Hayashida, M. 1989. Seed dispersal and regeneration pat- 
terns of Pinus parviflora var. pentaphylla on Mt. Apoi 
in Hokkaido. Research Bulletins of the College Experi- 
mental Forests, Faculty of Agriculture, Hokkaido Uni- 
versity. 46-1: 177-190. 

Kajimoto, T. 1989. Above-groimd biomass and litterfall of 
Pinus pumila scrubs growing on the Kiso Mountain 
Range in central Japan. Ecological Research. 4: 55-69. 

Koiziuni, T. 1974. Landschaftsokologische Unter- 
suchungen in der Alpine Stufe des Kisokomagatake in 
der Japanischen Zentralalpen mit Besonderer Beriick- 
sichtigung der Vegeta Strukturbodens. Japanese Jour- 
nal of Ecology. 24(2): 78-91. 

Mizoguchi, T.; Okitsu, S. 1987. Cone production of Pinus 
pumila Regal in high mountains in central Japan with 
special attention to basal diameter and annual elonga- 
tion on it. In: Transactions of the 98th annual meeting 
of the Japanese Forestry Society; 1987 April; K5aishu: 
363-364. 

Morgan, P.; Bunting, S. C. 1992. Technical note: cone 
crops of whitebark pine. Western Journal of Applied 
Forestry. 7(3): 71-73. 

Nakashinden, I. 1990. Pine cone production of Pinus 
pumila and their environment in Mt. Kisokomagatake, 
Nagano prefecture. In: Transactions of the 38th autum- 
nal meeting of the Association of Japanese Geogra- 
phers; 1990 September; Niigata: 104-105. 

Nakashinden, I. 1991. Pine cone production of Pinus 
pumila reconstructed from pine cone scars. In: Transac- 
tions of the 39th spring meeting of the Association of 
Japanese Geographers; 1991 April; Tokyo: 28-29. 

Okitsu, S. 1984b. Comparative studies on the Japanese 
alpine zone, with special reference to the ecology of 
Pinus pumila thickets. Geographical Review of Japan. 
57(11): 170-172. 

Okitsu, S. 1987. Pinus pumila zone. In: Ito, K., comp. Veg- 
etation of Hokkaido. Hokkaido, Japan: Hokkaido Univer- 
sity Press: 129-167. 



191 



Okitsu, S.; Ito, K. 1983. Dynamic ecology of the Pinus 
pumila community of Mts. Taisetsu, Hokkaido, Japan. 
The Environmental Science, Hokkaido University. 6(1): 
151-184. 

Okitsu, S.; Mizoguchi, T. 1986. Cone production of Pinus 
pumila community on Mt. Senjo, Akaishi Range, cen- 
tral Japan. In: Transactions of the 97th annual meeting 
of the Japanese Forestry Society; 1986 April; 
Utsunomiya: 343-344. 

Okitsu, S.; Mizoguchi T. 1990. Relation between cone 
production and stem diameter and stem elongation 
of Pinus pumila of Japanese high mountains. Japanese 
Journal of Ecology. 40(2): 49-55. 

Okitsu, S.; Mizoguchi, T. 1991. Cone production of Pinus 
pumila scrubs. Japanese Journal of Ecology. 41(2): 
101-107. 

Saito, S.; Kawabe, M.; Nakaoka, T. 1989. On the forest 
vegetation of Mt. Upepesanke, Tokachi, Hokkaido. (1): 
On a thicket of Pinus pumila at the ridge of 1610 m. 



Bulletin of the Higashi Taisetsu Museum of Natural 
History. 11: 23-34. 
Saito, S.; Kawabe, M. 1990. On the forest vegetation of 
Mt. Higashi-Nupukaushinupuri, Tokachi, Hokkaido. 
(2): On two thickets of Pinus pumila. Bulletin of the 
Higashi Taisetsu Museimi of Natural History. 12: 
17-29. 

Tomback, D.; Linhart, Y. 1990. The evolution of bird- 
dispersed pines. Evolutionary Ecology. 4: 185-219. 

Weaver, T.; Forcella, F. 1986. Cone production in 
Pinus albicaulis forests. In: Shearer, R. C, comp. 
Proceedings — conifer tree seed in the Inland Mountain 
West; 1985 August 5-6; Missoula, MT. Gen. Tech. Rep. 
INT-203. Ogden, UT: U.S. Department of Agriculture, 
Forest Service, Intermountain Research Station: 68-76. 

Yanagimachi, O.; Ohmori, H. 1991. Ecological status of 
Pinus pumila scrub and the lower boimdary of the 
Japanese alpine zone. Arctic and Alpine Research. 
23(4): 242-435. 



192 



EFFECTS OF SEED DISPERSAL BY 
CLARK'S NUTCRACKER ON EARLY 
POSTFIRE REGENERATION OF 
WHITEBARK PINE 

Diana F. Tomback 



Abstract — Clark's nutcrackers {Nucifraga columbiana) bury 
whitebark pine (Pinus alhicaulis) seeds in recently burned ter- 
rain. Studies of early postfire whitebark pine regeneration in the 
Northern Rocky Mountains indicate that seed dispersal by nut- 
crackers (1) may be more effective than seed dispersal by wind in 
restocking large bums, particularly when the seed source is unfa- 
vorably positioned with respect to prevailing winds, and (2) pro- 
duces a negative exponential distribution when whitebark pine 
regeneration density is plotted vs. distance from the seed source. 
Whitebark pine seeds can remain dormant for one or more years 
after dispersal, which results in continuity of regeneration over 
time. 



1986), although updrafts and storms may disperse some 
seeds over longer distances. Often, the regeneration from 
nutcracker caches is in the form of whitebark pine seed- 
Ung clusters (for example see Tomback 1982, 1986; Tomback 
and others 1990, 1993), which may result in mature multi- 
genet tree clusters (see Tomback and Schuster, these pro- 
ceedings, and references therein). In this paper I summa- 
rize the results of recent studies that compare patterns of 
early postfire regeneration of the bird-dispersed white- 
bark pine and its wind-dispersed forest associates. 

REGENERATION IN WESTERN 
MONTANA 

Studies of two large, severe subalpine bums in western 
Montana (Bitterroot National Forest, Ravalli County) 
illustrate how seed dispersal by nutcrackers affects pat- 
terns of early postfire regeneration of whitebark pine and 
associated conifers. These two study areas were selected 
for similarities in topography, severity of bum, forest 
types, and time elapsed since fire. In this region of west- 
em Montana, whitebark pine is abundant between about 
2,290 and 2,620 m elevation (Amo and Hoff 1989; Pfister 
and others 1977), and the prevailing winds in late sum- 
mer and fall are from the west and southwest (Finklin 
1983). The prefire forest in both study areas was a serai 
whitebark pine community. Reported results fi-om these 
studies are summarized here from Tomback and others 
(1990, 1993). 

The Sleeping Child Bum resulted from a wildfire in 
1961 that bumed about 11,350 ha of forest on the west 
slope of the Sapphire Range. The nearest whitebark pine 
seed sovu*ce is mature forest above 2,250 m on the eastern 
edge of the bum. Subalpine fir, Engelmann spmce (Picea 
engelmannii), and lodgepole pine (Pinus contorta) are also 
present in this upper seed source as well as around the 
lower perimeters of the bum. Because the prevailing 
winds are fi*om the west, vdnd-dispersed seeds are irregu- 
larly dispersed into the eastem portion of the bum fi'om 
the upper seed source. 

In 1987, we established a 3.6-km-long transect along 
an eastwest trending ridge, running from the seed soiirce 
into the northem center of the burn (fig. 2 in Tomback 
and others 1990). Elevation along the ridge decreases 
from about 2,480 to 2,150 m with increasing distance 
from the eastern edge of the bum. Every 150 m along the 
transect we established parallel quadrats, each 50 m long, 
on the ridgetop, north, and south aspects of the ridge for 



Fire is an ecologically important, recurring event in 
most western North American montane forests, including 
those regions where whitebark pine {Pinus albicaulis) is 
found. Fire-prone areas are those with fuel accumulation, 
dry, windy weather, and periods of frequent lightning 
storms. In the Northem Rocky Mountains, fire burns 
whitebark pine communities every 50 to 300 years (Amo 
1980; Amo and Petersen 1983; Romme 1980). Stand- 
replacing fires result in the renewal of successional white- 
bark pine communities, and underbums kill the less fire- 
resistant subalpine fir {Abies lasiocarpa). In the absence 
of fire in these commimities, whitebark pine is often re- 
placed by more shade-tolerant conifers, particularly sub- 
alpine fir (Amo 1986; Amo and Hoff 1989). 

Studies of early postfire regeneration, defined here as 
the first 3 decades after fire, provide information on the 
dynamics of seed dispersal and seedling establishment for 
whitebark pine and associated conifers. Clark's nutcrack- 
ers {Nucifraga columbiana) are the principal dispersal 
agents for whitebark pine seeds, in contrast to other coni- 
fers whose seeds are wind-dispersed. Nutcrackers bury 
caches of one to 15 or more ripe whitebark pine seeds at 
distances from parent trees of a few meters to 22 km 
(Hutchins and Lanner 1982; Tomback 1978, 1982; Vander 
Wall and Balda 1977; see review in Tomback and Linhart 
1990). In contrast, most vraid-dispersed seeds land only 
within about 120 m of parent trees (McCaughey and others 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Diana F. Tomback is Associate Professor of Ecology, Department of 
Biology and Center for Environmental Sciences, University of Colorado 
at Denver, Denver, CO 80217-3364. 



193 



a total of 63 quadrats. Quadrat widths varied in preset 
increments from 1.25 to 15 m, with respect to local densi- 
ties of whitebark pine. For each quadrat we recorded all 
conifer regeneration. 

The Saddle Moimtain Burn resulted from a wildfire in 
1960 that spread over 1,240 ha of forest in the Bitterroot 
Mountains west of Lost Trail Pass. Running northeast 
to southwest in the longest dimension, the burned area 
ranges in elevation from about 1,950 to 2,475 m. On the 
southwest edge of the burn above 2,275 m, the whitebark 
pine seed source is positioned so that the burned area is 
leeward with respect to prevailing winds. Subalpine fir, 
Engelmann spruce, and lodgepole pine also occur in this 
upper seed source and, with Douglas-fir {Pseudotsuga 
menziesii), around the lower perimeters of the burn. 

In 1988 we established two parallel 3.2-km-long tran- 
sects beginning at the seed source and heading northeast 
through the southern portion of the burn (fig. 4 in Tomback 
and others 1990). We set up 41 quadrats and recorded all 
regeneration according to the general methodology out- 
lined earlier for the Sleeping Child Burn (for details of 
methodology, see Tomback and others 1990, 1993). 

Mean whitebark pine regeneration density sampled per 
quadrat in the Sleeping Child Burn was much greater 
than that of its shade-tolerant competitors, subalpine fir 



Table 1 — Mean density and standard deviation (S.D.) per quadrat 
for conifer^ regeneration 26 years and 28 years after fire 
in tlie Sleeping Child and Saddle Mountain Burns, 
respectively. Based on Tomback and others (1 990) 

Regeneration sites per 
ABLA PIEN PICO PSME 



PIAL 



Sleeping Child 

X 0.070 0.008 0.004 0.039 

S.D. .104 .014 .009 .066 

Saddle Mountain 

X .044 .046 .030 .323 

S.D. .050 .068 .044 .355 



0.019 
.032 



TIAL = whitebark pine, ABLA = subalpine fir, PIEN 
PSME = Douglas-fir. 



Engelmann spruce. 



and Engelmann spruce, and greater than the shade- 
intolerant lodgepole pine as well (table 1). Scatterplots of 
quadrat densities vs. distance from the upper seed source 
revealed interesting differences in regeneration patterns 
among the conifers (fig. 1). The scatterplot for whitebark 
pine was a negative exponential curve, with the highest 
densities near the seed source and a long tail indicating 



0.512 



0.400 - 



0.200 - 

0 



'3) 
c 

Q 



p. albicaulis 



• • • • • 
• • • • • 

.• • . J J I It •• • • 



0.296 |- p. contorta 



0.200 



0.100 



• • • • 

- • • • • 



• • • 



0.08 r A. lasiocarpa 



0.05 



0.02 



0.056 r p engelmannii 



0.040 



• • • 



0.020 



• I 



• I • 
• • • 



I I »» w I l,« « i »» . I » y I « I I 



1000 



2000 



3000 3650 



1000 



2000 



3000 3650 



Distance (m) 



Figure 1 — Scatterplots of regeneration density per quadrat versus distance from the whitebark pine 
seed source for four conifer species in the Sleeping Child Burn. Clockwise from left: whitebark pine, 
lodgepole pine, Engelmann spruce, subalpine fir. From Tomback and others (1990). 



194 



some regeneration beyond 3.6 km. Quadrat density vs. 
distance in subalpine fir showed a negative linear rela- 
tionship, indicating that the primary seed soiirce for this 
species was also the upper, unbumed forest. In contrast, 
the scatterplot for Engelmann spruce was nearly flat, 
with the highest values at the lower end of the transect, 
suggesting that seeds may be blown into the ridge area 
from both the higher and lower seed sources. For lodge- 
pole pine, densities increased with distance from the up- 
per, iinbumed forest (are higher at lower elevations). 
Before the fire, the lower and middle elevations of this 
area were dominated by lodgepole pine, and thus much of 
the postfire regeneration may have come from seeds from 
serotinous cones within the bum (Lotan 1976). With the 
exception of Engelmann spruce, the density vs. distance 
relationships were statistically significant atP < 0.01 or 
greater (Tomback and others 1990). 

For the Saddle Moimtain Burn, mean whitebark pine 
regeneration density was not significantly different than 
that for the Sleeping Child Bum, although somewhat 
lower (table 1). The density was actually comparable to 
that obtained for the ridgetop quadrats (west facing) in 
the Sleeping Child Bum. In contrast, the densities for 
subalpine fir and Engelmann spruce were much greater 
than those in the Sleeping Child Burn. Lodgepole pine 
was particularly abimdant as well. Scatterplots of quad- 
rat density vs. distance from the upper seed source were 
comparable to those from Sleeping Child (fig. 2). 



0.14 



0.10 



r •• 



(0 

c 

0) 
Q 



0.05 

0 

0.30 
0.20 

0.10 



P. albicaulis 



• • • 



A. lasiocarpa 



•• • 



• • • • 

• ■ • • • • 

• , • •••••• • 



1000 



2000 



3000 



Distance (m) 

Figure 2 — Scatterplots of regeneration density 
per quadrat versus distance from the whitebark 
pine seed source for whitebark pine (top) and 
subalpine fir (bottom) in the Saddle Mountain 
Burn. From Tomback and others (1990). 



Table 2 — Maximum age differences within clusters of whitebark 

pine seedlings. Cluster sample size for the Sleeping Child 
Burn = 87; for the Saddle Mountain Burn = 7. Based on 
table 4, Tomback and others (1993) 



Maximum age 
differences 


Sleeping Child 


Saddle Mountain 


Years 


n 


Percent 


n 


Percent 


0 


30 


34.5 


1 


14.3 


1 


24 


27.6 


4 


57.1 


2 


12 


13.8 


1 


14.3 


3 


10 


11.5 


0 


0 


4 


6 


6.9 


1 


14.3 


5 


1 


1.2 






6 


1 


1.2 






7 


2 


2.3 






8 


1 


1.2 







Whitebark pine regeneration showed a negative exponen- 
tial distribution, and subalpine fir a similar but more lin- 
ear negative relationship between quadrat density and 
distance from the upper, unbiu-ned forest (density vs. 
distance relationships significant at P < 0.001 for both). 
Similar to findings for the Sleeping Child Bum, the En- 
gelmann spruce scatterplot was nearly flat. For both 
Douglas-fir and lodgepole pine, density increased with 
distance from the upper, imbumed forest; major seed 
sources for both species were at the lower edges of the 
bum and, in the case of lodgepole pine, seeds from within 
the bum. However, the relationship between density and 
distance was not statistically significant for these latter 
three conifers. 

The distinctive negative exponential shape of the white- 
bark pine regeneration curve probably results from two 
factors: (1) whitebark pine seeds are only available at the 
upper elevations of these burns, and (2) nutcrackers tend 
to bury higher densities of seeds near parent trees. Ap- 
parently, nutcrackers do cache seeds at long distances 
from parent whitebark pine trees, because we found low 
densities of regeneration up to 8 km from the whitebark 
pine seed source (Tomback and others 1990). It could be 
argued that the concomitant decrease in elevation in both 
study areas was a confounding variable in the whitebark 
pine density vs. distance relationship. However, recently, 
I tested this possibility by examining whitebark pine re- 
generation along a 3.4-km transect following an eleva- 
tional isocline in the Sundance Bum of northern Idaho. 
Like the Sleeping Child and Saddle Mountain Bxirns, the 
whitebark pine seed source was restricted to one side of 
the bum, and the resulting regeneration curve followed 
a negative exponential distribution (Tomback 1992). 

It appears that seed dispersal by nutcrackers has 
provided whitebark pine an advantage over its shade- 
tolerant competitor, subalpine fir, in the upper subalpine 
of the Sleeping Child Burn. Because wind-dispersed fir 
seeds from the upper seed source must travel against the 
prevailing winds and cover a long distance, the regenera- 
tion density of fir is lower than that for whitebark pine. 
In contrast, in the smaller Saddle Mountain Burn, white- 
bark pine and subalpine fir have comparable regeneration 



195 



densities. In this situation, dispersal of subalpine fir seeds 
(and Engelmann spruce seeds) is favored both by the pre- 
vailing winds from the west and the much smaller size of 
the burn. Thus, for restocking burns, seed dispersal by 
nutcrackers may provide whitebark pine an advantage 
over wind-dispersed conifers when (1) the size of the burn 
is large compared to distances traveled by wind-blown 
seeds, (2) the seed source is unfavorably positioned with 
respect to prevailing winds, and (3) the seed source is 1 km 
or more from the perimeter of the burn. 

Another consequence of seed dispersal by nutcrackers 
is that seedling or sapling clusters occurred at more than 
40 percent of the whitebark pine regeneration sites in both 
the Sleeping Child and Saddle Mountain Burns (Tomback 
and others 1990, 1993). Cluster sizes in the Sleeping 
Child Burn ranged from two to eight stems with an over- 
all mean of 1.9 stems per regeneration site (n = 455 total 
sites). In the Saddle Mountain Burn, clusters ranged 
from two to 10 stems, with an overall mean of 2.0 stems 
per regeneration site (n = 164 sites). We aged two or more 
cluster members from a cluster subsample from each of 
the study areas (Tomback and others 1993). For the 
Sleeping Child Burn, there were no differences in age 
among seedlings within a cluster for 34 percent of the 
clusters, differences of 1 year for 28 percent of the clus- 
ters, and differences of 2 to 8 years within clusters for the 
rest of the subsample (total n = 87 clusters; all seedlings 
aged in 63 clusters) (table 2). Of the small number of 
clusters sampled in the Saddle Mountain Burn, five clus- 
ters (71 percent) had age differences of 0 to 1 year, and 
the remaining two clusters had seedlings differing by 
2 and 4 years, respectively (table 2). These resiilts strongly 
suggest that asynchronous germination of seeds may oc- 
cur in some clusters. This could be explained by delayed 
maturation of some seeds or different sensitivities to envi- 
ronmental conditions. Laboratory (Leadem 1986; Pitel and 
Wang 1990) and experimental field studies (McCaughey 
1993) indicate that embryo underdevelopment and de- 
layed germination are tjrpical in whitebark pine seeds. 
Delayed germination of seed caches would result in seed- 
ling establishment following little or no cone production 
and, thus, continuity of regeneration over time. 

We compared age structiire (year of tree establishment 
vs. time since fire) of regeneration for whitebark pine 
and subalpine fir in both the Sleeping Child and Saddle 
Mountain Burns to determine any important differences 
(Tomback and others 1993). Both species appeared to be 
similarly affected by environmental and ecological factors: 
a disproportionately high percentage of regeneration was 
established between 1977 and 1985, about 17 to 25 years 
after fire (see fig. 1 in Tomback and others 1993). Thus, 
major tree recruitment was episodic. A small increase 
in mean January through August precipitation between 
1977 and 1985 possibly contributed to this trend. If we 
can generalize from these results, it appears that white- 
bark pine and subalpine fir become established synchro- 
nously after fire, albeit at different densities depending 
on relative seed dispersal abilities. 



REGENERATION IN THE GREATER 
YELLOWSTONE AREA 

Studies in the Greater Yellowstone Area following the 
major fires of 1988 have provided comparative informa- 
tion on patterns of seed dispersal, germination, and seed- 
ling mortality in serai whitebark pine communities 
(Tomback 1991, 1993). Unlike the large-scale, stand- 
replacing burns described earlier for western Montana, 
the Yellowstone fires created patches of burned areas of 
different severities, often surrounded by unburned forest. 

In 1990, my research assistants and I established a to- 
tal of 275 permanent plots, each 20 m^, in two different 
study areas. In the Cooke City study area, Gallatin Na- 
tional Forest, we placed plots in four sample sites, repre- 
senting the following treatments: dry, severely biirned 
(50 plots); moist, severely burned (50 plots); dry, imbumed 
(25 plots); moist, unburned (25 plots). Plot aspects in this 
area ranged from 80° to 180°, and elevations rsmged from 
2,680 to 2,745 m. On Mount Washbiirn, Yellowstone 
National Park, we selected three sample sites, represent- 
ing the following treatments: dry, severely burned (50 
plots); moist, severely burned (50 plots); moist, moder- 
ately burned (25). Aspects ranged from 250° to 360°, 
and elevations ranged from 2,560 to 2,745 m. All conifer 
regeneration was recorded for these plots in the siunmers 
of 1990, 1991, and 1992; whitebark pine regeneration was 
mapped for each plot. I briefly summarize the results 
here (Tomback 1991, 1993). 

Whitebark Regeneration Trends 

Despite an abundant whitebark pine cone crop in 1989, 
no whitebark pine regeneration occurred on any severely 
burned or moderately burned plots in 1990 in either study 
area (table 3). Only one plot in the moist, unburned treat- 
ment had one newly germinated seedling cluster. Al- 
though there was virtually no cone production in 1990, 
all severely burned and moderately burned treatments 
had some new whitebark pine regeneration in 1991, which 
was probably the result of delayed germination of seeds 
cached in 1989. Of the unburned treatments, again only 
the moist plots experienced any new regeneration (table 3). 

In 1992 there was some new whitebark pine regenera- 
tion on the Cooke City moist, severely burned plots and 
on the Mount Washburn moist, moderately burned plots. 
By 1992, the latter sample site supported the highest den- 
sity of whitebark pine regeneration of all sites (table 3). 
Seedling survivorship from 1991 to 1992 was fairly high. 
Those that survived to the summer of 1992 tended to be 
shaded by other plants or forest debris for at least part 
of the day. One cluster of seedlings that germinated in 
1991 just outside a plot on the Mount Washburn moist, 
severely burned sample site contained two new germinants 
in 1992. This observation confirms the possibility that 
seeds within caches may germinate in different years, as 
illustrated by the ages of cluster members from western 
Montana. 



196 



Table 3 — Whitebark pine regeneration densities for different plot 

treatments in the Cooke City and Mount Washburn study 
areas following the 1 988 Yellowstone fires 



Mean density of regeneration sites^ per 
Treatment 1990 1991 1992 



Cooke City 



Dry, severely burned 


0 


0.012 


0.015 


Moist, severely burned 


0 


.014 


.010 


Dry, unburned 


0 


0 


.002 


Moist, unburned 


0.002 


.010 


.008 


Mount Washburn 








Dry, severely burned 


0 


.010 


.011 


Moist, severely burned 


0 


.021 


.016 


Moist, moderately burned 


0 


.030 


.038 



'A single regeneration site may support either a single seedling or a cluster 
of seedlings arising from one seed cache. 



Regeneration Trends of Other 
Conifers 

Trends in the regeneration of other conifers differed be- 
tween the Cooke City and Mount Washburn study areas, 
probably as a result of seed availability and seedbed con- 
ditions. For example, no conifer regeneration, old or new, 
occiirred on the Cooke City dry, severely burned plots in 
1990; small niombers of Engelmann spruce and subalpine 
fir occurred on these plots in subsequent years. On the 
moist, severely burned plots, new spruce germinants 
dominated in 1990 and both spruce and fir, new and old, 
in subsequent years. Primarily new fir regeneration oc- 
curred each year on the Cooke City dry, unburned plots, 
with few older firs surviving; and, mostly spruce occurred 
on the moist, unburned plots. Thus, spruce and fir regen- 
eration predominated in the Cooke City area. Few lodge- 
pole pine were apparent in the prefire forest or in the un- 
bumed forest around the treatment sites. 

In the Moimt Washburn study area, old and new lodge- 
pole pine and new Engelmann spruce occiured in 1990 on 
the dry, severely biirned plots. Older lodgepole pine and 
some new spruce occurred in subsequent years. On the 
moist, severely burned plots, new lodgepole pine and 
spruce regeneration predominated in 1990, and older 
lodgepole pine and spruce in 1991. In 1992 there were 
munerous new germinants of both species and some sub- 
alpine fir. High numbers of new spruce seedlings domi- 
nated the moist, moderately burned plots in 1990, and 
older and new spruce in subsequent years. Thus, lodge- 
pole pine and spruce regeneration predominated in the 
burned area sampled on Moimt Washbiirn. Lodgepole 
pine had been an important species in the prefire forest 
and was present, along with spruce, in imburned forest 
not far fi-om the plot treatment sites. The high spruce re- 
generation density on the moderately burned plots may 
have been the consequence of a more favorable seedbed. 

Although the numbers of whitebark pine germinants 
were lower in most cases than those of the other conifer 
species in both study areas, their siu-vivorship from 1991 



to 1992 was relatively high. The microsites selected for 
seed caching by nutcrackers and the hardiness of the 
whitebark pine seedlings may result in lower mortality 
compared to other conifers. We must follow the regenera- 
tion patterns of whitebark pine and its forest associates 
for several more years to test this idea. 

CONCLUSIONS 

The patterns of whitebark pine regeneration that we 
observe are the consequence of the seed-dispersal behav- 
iors and site preferences of nutcrackers, as well as their 
tendency to bury multiseed caches. Nutcrackers bury 
seeds in burned terrain soon after fire (for example, see 
Tomback and Knowles 1989). Because whitebark pine 
seeds vary in maturation rate, nutcrackers cache both de- 
veloped and imderdeveloped seeds (Leadem 1986), result- 
ing in germination over several different years from the 
same seed crop. 

Early postfire regeneration of whitebark pine may be the 
consequence of rapid and widespread seed dispersal by 
nutcrackers, suitability of the cache sites selected by nut- 
crackers, and the hardiness of the seedlings themselves. 

ACKNOWLEDGMENTS 

The Intermountain Research Station, USDA Forest Ser- 
vice, Ogden, UT, has funded all whitebark pine postfire 
regeneration studies in western Montana, northern Idaho, 
and the Greater Yellowstone Area. I thank Stephen Arno, 
Jim Brown, and Bob Keane of the Intermountain Re- 
search Station Fire Sciences Laboratory, Missoula, MT, 
for logistic support and many helpful discussions concern- 
ing this research. Steve Arno and Bob Keane reviewed 
the draft manuscript. Lyn Hoffinann and Sharren Reuss 
(Sund) provided assistance during the western Montana 
studies, Kathy Carsey and Mary Powell worked with me 
during the three field seasons of the Yellowstone study, 
and Mary Powell, Jane Kees, and Mark Flower were field 
assistants for the Idaho work. Support for attending the 
SSPE Workshop was provided by the College of Liberal 
Arts and Sciences, University of Colorado at Denver. 

REFERENCES 

Arno, S. F. 1980. Forest fire history in the northern 
Rockies. Journal of Forestry. 78(8): 460-465. 

Arno, S. F. 1986. Whitebark pine cone crops — a diminish- 
ing source of vdldlife food? Western Journal of Applied 
Forestry. 1(3): 92-94. 

Arno, S. F.; Petersen, T. D. 1983. Variation in estimates 
of fire intervals: a closer look at fire history on the Bit- 
terroot National Forest. Res. Pap. INT-301. Ogden, UT: 
U.S. Department of Agricult;ire, Forest Service, Inter- 
mountain Forest and Range Experiment Station. 8 p. 

Arno, S. F.; Hofif, R. J. 1989. Silvics of whitebark pine 
{Pinus albicaulis). Gen. Tech. Rep. INT-253. Ogden, 
UT: U.S. Department of Agriculture, Forest Service, 
Intermountain Research Station, lip. 



197 



Finklin, A. I. 1983. Weather and climate of the Selway- 
Bitterroot Wilderness. Moscow, ID: University Press 
of Idaho. 144 p. 

Hutchins, H. E.; Lanner, R. M. 1982. The central role of 
Clark's nutcracker in the dispersal and establishment 
of whitebark pine. Oecologia. 55: 192-201. 

Leadem, C. L. 1986. Seed dormancy in three Pinus species 
of the Inland Mountain West. In: Shearer, R. C, comp. 
Proceedings — conifer tree seed in the Inland Mountain 
West symposium; 1985 August 5-6; Missoula, MT. Gen. 
Tech. Rep. INT-203. Ogden, UT: U.S. Department of 
Agrictdture, Forest Service, Intermountain Research 
Station: 117-124. 

Lotan, J. E. 1976. Cone serotiny — fire relationships in 
lodgepole pine. In: Lyon, L. J., comp. Tall Timbers Fire 
Ecology Conference No. 14 and Intermountain Fire Re- 
search Council fire and land management symposium: 
proceedings; Missoula, MT: 267-278. 

McCaughey, W. W.; Schmidt, W. C; Shearer, R. C. 1986. 
Seed dispersal characteristics of conifers in the Inland 
Mountain West. In: Shearer, R. C, comp. Proceedings — 
conifer tree seed in the Inland Mountain West sympo- 
sium; 1985 August 5-6; Missoula, MT. Gen. Tech. Rep. 
INT-203. Ogden, UT: U.S. Department of Agriculture, 
Forest Service, Intermountain Research Station: 50-62. 

McCaughey, W. W. 1993. Delayed germination and seed- 
ling emergence of Pinus albicaulis in a high elevation 
clearcut in Montana, USA. In: Edwards, D. G., comp. 
Seed dormancy and barriers to germination: proceed- 
ings lUFRO symposium; 1991 April 23-26; Victoria, BC. 
Victoria, BC: Forestry Canada. [In press]. 

Pfister, R. D.; Kovalchik, B. L.; Arno, S. F.; Presby, R. C. 
1977. Forest habitat types of Montana. Gen. Tech. Rep. 
INT-34. Ogden, UT: U.S. Department of Agriculture, 
Forest Service, Intermountain Forest and Range Ex- 
periment Station. 174 p. 

Pitel, J. A.; Wang, B. S. P. 1990. Physical and chemical 
treatments to improve germination of whitebark pine 
seeds. In: Schmidt, W. C; McDonald, K. J., comps. 
Proceedings — symposium on whitebark pine ecosys- 
tems: ecology and management of a high-mountain 
resource; 1989 March 29-31; Bozeman, MT. Gen. Tech. 
Rep. INT-270. Ogden, UT: U.S. Department of Agricul- 
ture, Forest Service, Intermountain Research Station: 
130-133. 

Romme, W. H. 1980. Fire frequency in subalpine forests 
of Yellowstone National Park. In: Fire history work- 
shop: proceedings; 1984 October 20-24; Tucson, AZ. 



Gen. Tech. Rep. RM-81. Fort Collins, CO: U.S. Depart- 
ment of Agriculture, Forest Service, Rocky Mountain 
Forest and Range Experiment Station: 27-30. 

Tomback, D. F. 1978. Foraging strategies of Clark's nut- 
cracker. The Living Bird. 16: 123-161. 

Tomback, D. F. 1982. Dispersal of whitebark pine seeds 
by Clark's nutcracker: a mutualism hypothesis. Journal 
of Animal Ecology. 51: 451-467. 

Tomback, D. F. 1991. Progress report on 1991 fieldwork 
to U.S. Department of Agriculture, Forest Service, In- 
termountain Research Station, Ogden, UT: patterns of 
postfire regeneration of whitebark pine in the Greater 
Yellowstone Area. 

Tomback, D. F. 1992. Patterns of postfire whitebark pine 
regeneration in the Sundance Burn of northern Idaho. 
Unpublished data on file at: Department of Biology, 
University of Colorado at Denver, Denver, CO. 

Tomback, D. F. 1993. Progress report on 1992 fieldwork 
to U.S. Department of Agriculture, Forest Service, In- 
termountain Research Station, Ogden, UT: patterns of 
postfire regeneration of whitebark pine in the Greater 
Yellowstone Area. 

Tomback, D. F.; Knowles, J. W. 1989. Observations on 
nutcracker seed harvesting and caching on Mount 
Washburn following the 1988 Yellowstone fires. Unpub- 
lished data on file at: Department of Biology, University 
of Colorado at Denver, Denver, CO. 

Tomback, D. F.; Linhart, Y. B. 1990. The evolution of bird- 
dispersed pines. Evolutionary Ecology. 4: 185-219. 

Tomback, D. F.; Hoffmann, L. A.; Sund, S. K. 1990. Co- 
evolution of nutcrackers and pines: implications for 
forest regeneration. In: Schmidt, W. C; McDonald, 
K. J., comps. Proceedings — symposium on whitebark 
pine ecosystems: ecology and management of a high- 
mountain resource; 1989 March 29-31; Bozeman, MT. 
Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Department 
of Agriculture, Forest Service, Intermountain Research 
Station: 118-129. 

Tomback, D. F.; Sund, S. K.; Hoffmann, L. A. 1993. Post- 
fire regeneration of Pinus albicaulis: height-age rela- 
tionships, age structure, and microsite characteristics. 
Canadian Journal of Forest Research. 23: 113-119. 

Vander Wall, S. B.; Balda, R. P. 1977. Coadaptations of 
the Clark's nutcracker and the pinon pine for efficient 
seed harvest and dispersal. Ecological Monographs. 47: 
89-111. 



198 



NEW TREND IN DENDROCHRONOLOGY: 
1. THEORETICAL PRINCIPLES OF 
REPROCHRONOLOGY 

Vladislav N. Vorobjev 



Abstract — Reviews previous research pertinent to tree growth 
and cone-bearing relationships in Siberian stone pine (Pinus 
sibirica) and other species. The ultimate objective is to secure 
stable and moderately high cone production without significantly 
jeopardizing tree growth. To reach this objective the author 
points out potential avenues of research in tree growth dynamics 
and reproductive activities, utilizing procedures used in 
dendrochronology. 



Dendrochronology was established as a science of the 
growth dynamics of trees, the structure of the annual 
ring, and the effect of internal and, especially, external 
factors. The latter determined the development of den- 
drochronology and a number of other special scientific 
areas. 

Generative processes have been studied separately and 
dealt with by the corresponding branch of seed farming in 
terms of the concept of the crop dynamics or crop cyclicity. 
In doing so, novel independent results have been obtained 
and, more important, premises have been created for com- 
bining these findings with dendrochronology. 

The correlation between the growth and generative de- 
velopment of higher plants comprises a number of aspects 
and has been studied for a long time. The study of herba- 
ceous plants has lagged behind that of tree investigations. 
Suffice it to say that not only have monographs on this 
subject been lacking, but original publications contribut- 
ing to this field have also been scarce. The only exception, 
to the best of oiu- knowledge, has been the work done 
in Tomsk by the laboratory of cone-bearing trees 
(Goroshkevich 1989; Nekrasova 1974; Vorobjev 1983; 
Vorobjev and Vorobjeva 1982; Vorobjev and others 1989). 
However, the problem of the relation between growth and 
cone bearing (in coniferous trees, in particular) is of great 
importance in view of the necessity to increase cone crops. 
While this problem is being solved in herbaceous plants 
and fi-uit-bearing trees by an intense short-term shift of 
metabolism toward activization of the generative pro- 
cesses, this technique fails in coniferous trees because 
of the long growth period inherent in them. 

It is assumed on the basis of the foregoing consider- 
ations that control of the sexual reproduction of trees 



Poster presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Vladislav N. Vorobjev is Head of the Academy of Sciences, Siberian 
Branch, Institute for Ecology of Natural Complexes, Academichesky pr. 2, 
Academgorodok, Tomsk, 634055, Russia. 



must be related to the determination of optimal correla- 
tions between the generative and growth processes 
through the long-term activization of the vital activity 
of plants. 

PREVIOUS RESULTS 

The expedience of this approach was also confirmed 
by the results of our previous investigations. The experi- 
ments on the resin tapping of the Siberian cedar pine, 
sometimes called Siberian stone pine {Pinus sibirica), 
showed that the increase of cone crops by creating favor- 
able conditions for cone bearing alone had no futiu'e be- 
cause it would eventually resiilt in the weakening of 
the growth processes and then of the generative ones 
(Vorobjev 1974). Therefore, it appears necessary to look 
for certain correlations between growth and cone bearing 
that would make it possible to secure moderately high 
cone crops on a long-term stable basis. It is this problem 
that seems to be of great theoretical and practical interest 
(Vorobjev and Vorobjeva 1980). 

The study of the effect of the reproductive activity on 
shoot growth and the shoot condition on subsequent cone 
bearing provides insight into the relation between tree 
growth and cone bearing. Observations have shown that 
cone bearing depends on the previous shoot growth in 
Scotch pine (Pinus sylvestris), English oak (Quercus sp.), 
and apple (Malus sp.) trees (Minina 1954; Polozova 1957; 
Pravdin 1950; Tsel'niker and Semikhatova 1957). It has 
been pointed out that optimal shoot growth is required to 
induce cone bearing. Of no less importance is the time it 
takes to achieve growth, since the flower germs are initi- 
ated when apical growth gets weaker or terminates. The 
surface area of leaves also plays a significant role. 

The effect of cone bearing on the current increment has 
been the subject of a number of papers. In particular, the 
diameter increment in coniferous trees was shown to de- 
crease in the cone-maturing year (Chalupka and others 
1975, 1976, 1977; Danilov 1953). However, the effect of the 
decreased diameter increment on fiirther seminiference of 
trees and metamere growth has not been examined yet. 

The change in shoot growth due to cone bearing is the 
net effect of sexualization and subsequent development 
of generative organs at different stages of their maturity. 
The dependence of the cambial growth of the shoot on its 
sex has been described in literature (Nekrasova 1972; 
Varnell 1976). However, no experimental evidence of the 
role of sexualization and nutrition of the generative or- 
gans, and of the effect of the nutrition of the 1- or 2-year 
cones on the annual ring and its internal structure in the 
growing shoot has been available (Vorobjev and Vorobjeva 
1982). 



199 



RETROSPECTIVE METHOD 

Good possibilities for the study of the effect of cone 
crops on tree increment (both Hnear and cambial) have 
opened up due to recent development of the method of ret- 
rospective accoimt of the reproductive activity of trees 
during their lifetime (see Vorobjev and others, these pro- 
ceedings). Significantly, this procedure can also be used 
to reproduce male "flowering" and the formation of a num- 
ber of vegetative metameres, though for a shorter period 
of time. In this case, shoots, branches, and crown become 
major objects of inquiry in the investigation of the relation 
between the growth and generative development of trees. 

The research done into the growth-cone-bearing rela- 
tionship in Siberian cedar pine in West Siberia and Gomy 
Altai for the last 100 years or more (Vorobjev 1983) sug- 
gests a new trend in the investigations made at the cross- 
roads of the two sciences, biology of reproductive develop- 
ment and dendrochronology. 

This trend can be referred to as reprochronology, com- 
prising both the cyclic fluctuations of the reproductive ac- 
tivity and its relation to the cambial growth. Perhaps, the 
word is not a happy choice from the terminology point of 
view, but it generally implies sexual reproduction. 

The time series of reproductive activity studied show 
cycles of variable duration ranging from 3 to 28 years 
or more. A 10- to 11-year reproductive cycle is less pro- 
noimced than the growth cycle. The relation between the 
crop dynamics and solar activity is feebly marked, and 
seed cycles fall both at the peak and decay of the solar 
activity. The reproductive activity is indirectly related to 
the cambial growth and is delayed with respect to the lat- 
ter by the time interval determined by the life cycle of one 
generation and the growth aftereffect. The peak of the 
siminiference activity generally falls at the descending 
branch of the solar cycle and the period of the retarded 
shoot growth. An increase in cone crops is related to the 
previous solar activity and vigorous growth. 

Thus it appears worthwhile to study further the struc- 
ture of the time series of the reproductive activity and their 
relation to the dynamics of growth of shoots, branches, and 
trees, to evaluate the effect of the external factors using 
new methods, and to develop prediction models. Also of 
scientific importance are individual, interpopulational, 
geographical, ecological, and other aspects of the relation 
between the growth and the generative development of 
trees as well as between the seminiference and the annual 
ring of the shoot and tree. It should be noted that the 
foregoing potential research areas suggest the use of the 
techniques and procedures adopted in dendrochronology. 

REFERENCES 

Chalupka, W.; Giertych, M.; Krolikowski, Z. 1975. The 
effect of cone crops on growth in Norway spruce {Picea 
abies (L.) Karst.). Arbor. Kor. 20: 201-212. 



Chalupka, W.; Giertych, M.; Krolikowski, Z. 1976. The 
effect of cone crops in Scotch pine on tree diameter in- 
crement. Arbor. Kor. 21: 361-366. 

Chalupka, W.; Giertych, M.; Krolikowski, Z. 1977. Rela- 
tion between specific gravity of wood in Norway spruce 
(Picea abies (L.) Karst.): some growth parameters and 
cone yield. Arbor. Kor. 22: 205-212. 

Danilov, D. N. 1953. Effect of cone-bearing on tree-ring 
structure in common spruce (Picea excelsa). Bot. J. 
38(3): 367-377. [In Russian]. 

Goroshkevich, S. N. 1989. The dynamics of the micro- 
strobile initiation in the male shoots of the Siberian ce- 
dar pine as related to weather conditions. Ecol. 6: 33-39. 
[In Russian]. 

Minina, E. G. 1954. Biological principles of flowering and 
fruiting oak. Transactions of Inst, of Forest of USSR 
Acad, of Sci. 17: 5-97. [In Russian]. 

Nekrasova, T. P. 1972. Biological principles of semi- 
niference of Siberian cedar pine. Novosibirsk: Nauka. 
272 p. [In Russian]. 

Nekrasova, T. P. 1974. The cyclicity of the cone-bearing of 
Siberian cedar pine. In: Biology of seed reproduction of 
coniferous trees of West Siberia. Novosibirsk: Nauka: 
70-75. [In Russian]. 

Polozova, (L.) Y. 1957. Growth of oak shoots as related to 
fruiting processes. Letters of Inst, of Forest of USSR 
Acad, of Sci. 9: 9-17. [In Russian]. 

Pravdin, (L.) F. 1950. Sexual dimorphism in Scotch pine 
(Pinus sylvestris L.). Transactions of Inst, of Forest of 
USSR Acad, of Sci. 3: 190-201. [In Russian]. 

Tsel'niker, Y. L.; Semikhatova, O. A. 1957. On the rela- 
tion between vegetative and generative stages of shoot 
development in certain trees. Bot. J. 42(7): 1044-1054. 
[In Russian]. 

Vamell, R. J. 1976. Correlation between branch charac- 
teristics and production of female strobile in longleaf 
pine. Forest Science. 22(2): 159-161. 

Vorobjev, V. N. 1974. Growth and generative processes 
in Siberian cedar pine under metabolic disturbance. In: 
Biology of seed reproduction of coniferous trees of West 
Siberia. Novosibirsk: Nauka: 140-170. [In Russian]. 

Vorobjev, V. N. 1983. Biological principles of compre- 
hensive utilization of Siberian cedar pine forests. 
Novosibirsk: Nauka. 254 p. [In Russian]. 

Vorobjev, V. N.; Vorobjeva, N. A. 1980. On the problem of 
regulation of seed crops of coniferous trees. In: Selec- 
tion, genetics and seed farming of trees as the basis for 
forming of highly productive forests. Moscow: 391-395. 
[In Russian]. 

Vorobjev, V. N.; Vorobjeva, N. A.; Goroshkevich, S. N. 
1989. Growth and sex of Siberian cedar pine. 
Novosibirsk: Nauka. 167 p. [In Russian). 

Vorobjeva, N. A.; Vorobjev, V. N. 1982. The effect of fe- 
male generative organs on the shoot growth in the 
Siberian cedar pine. Lesovedeneye (For. Sci.) 1: 51-57. 
[In Russian]. 



200 



NEW TREND IN DENDROCHRONOLOGY: 
2. METHOD OF RETROSPECTIVE 
STUDY OF SEMINIFERENCE DYNAMICS 
IN PINACEAE 

V. N. Vorobjev 

S. N. Goroshkevich 

D. A. Savchuk 



Abstract — Introduces a new method of retrospective analysis of 
cone crop dynamics of Siberian stone pine {Pinus sihirica) for up 
to 100 years before the present. For about the first 20 years, cone 
abscissions are discernible on the exterior of the shoots and can 
be used to reconstruct cone crop production, but after this they 
are not identifiable. A method was developed utilizing cross- 
sectional saw cuts to examine the interior of the branch shoots. 
Traces of the abscissions of cones remain in the shoots and, con- 
sequently, cone crops can be reconstructed for virtually the entire 
life of the tree with the procedures described in this paper. This 
method may also be useful for other conifers. 



A retrospective study of the life activity of trees has al- 
ways been of great interest to scientists. Retrospective 
analysis has been widely practiced in dendrochronologic 
and dendroclimatic studies focusing on the investigation 
of tree rings whose structure and dimensions are impor- 
tant indicators of the life activity of the tree organism. 
The variations of tree increment observed are basically 
related to the effect of exogenic factors, and cHmatic con- 
ditions in particular. As to the endogenic factors and, pri- 
marily, the reproductive activity, their contribution is rec- 
ognized but not accounted for because of the difficulty of 
estimating cone crops for long periods of time. In fact, 
this aspect of activity of the tree organism is regarded 
as the so-called "white noise" in a multifactor dendrochro- 
nological analysis of the effect of different conditions. 

The early information on the aimual dynamics of cone 
crops has been obtained from direct counts of the number 
of generative organs for several years. This technique is 
fairly widely used even today. However, applications of 
this technique are Hmited due to the necessity of long- 
term observations. Naturally, scientists sought to develop 
a method of retrospective study of the cone crop dynamics. 
This method is based on the fact that the cones and fruits 
either remain on the branches after the seed abscission 
or fall out, leaving traces on the shoot bark. The seasonal 



Poster presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

V. N. Vorobjev is Head of and S. N. Goroshkevich and D. A. Savchuk are 
Members of the Academy of Sciences, Siberian Branch, Institute for Ecol- 
ogy of Natural Complexes, 2, Acadcemichesky pr., Academgorodok, Tomsk, 
634055, Russia. 



character of the shoot growth in trees in the temperate 
zone makes it possible to age each shoot and, conse- 
quently, its generative organs. 

The first information we are aware of on the tise of the 
retrospective method of studying the crop dynamics can 
be foimd in the work of A. Renvall (1912) for the Scotch 
pine and N. Nesterov (1914) for some leaf -bearing trees 
(oak, maple, hawthorn). The method was further devel- 
oped and elaborated in great detail by Z. Trofimova 
(1953), T. Nekrasova (1957), and P. Gorchakovsky (1958) 
for different coniferous trees. A most comprehensive de- 
scription of this method was given by A. Korchagin (1960). 

The advent of the retrospective method has facilitated 
significant progress in the investigation of crop dynamics. 
The latter can be reproduced for the past 10-15 years on 
the basis of the cone traces on the shoot bark. Special em- 
phasis has been placed on Siberian cedar pine (sometimes 
called Siberian stone pine) because of the great natural 
and economic importance of its seed crops (Iroshnikov 
1963; Vorobjev 1964, 1974). As a result, a cycUc charac- 
ter of the fluctuations in the crops for different years was 
revealed and the relations between the seminiference, 
weather conditions, and tree grovTth were examined. 
However, further advance of the investigations required 
that the methodology be improved. 

In our opinion, there are two very essential ways of im- 
proving the method. First, work out a special technique 
for reconstructing the stages of the development of each 
generation from the initiation to the complete mattiration 
of cones and to determine the number of cones that pre- 
maturely abscissed in different stages of their develop- 
ment. These data will make it possible to study the crop 
dynamics and, in particular, its effect on the tree and 
shoot growth and the annual ring structure. Second, ex- 
tend the period of the retrospective account of crops from 
10-15 years to 100 years or more. 

The object of the present paper is to introduce a new 
method of retrospective analysis of the seminiference dy- 
namics of Siberian cedar pine, making use of the forego- 
ing information. 

CONE DEVELOPMENT 

Cones in the Siberian cedar pine are initiated in late 
July or early August (Nekrasova 1972). The germs of the 
female cones hibernate in the form of undifferentiated 



201 



Figure 1 — Female cone traces on the shoot bark, x 6: 
(a) cone that abscissed before pollination, (b) cone that 
abscissed after pollination, (c) cone that abscissed during 
maturation, (d) mature cone. 



primordia surrounded by bud scales. The primordia are 
located in the distal part of the winter terminal bud. The 
development of the female cones begins in May. An active 
linear growth of shoots occurs in the first half of June, 
and the cones resemble large lateral buds in the distal 
part of the growing shoot. Special observations show that 
the development of some of the female cones ceases at this 
period of time and they abort. For example, normally de- 
veloping cones were bright green and 10-15 mm long in a 
Siberian cedar pine forest near a settlement in the south 
of the Tomsk region on J\me 6, 1990. Some of the female 
buds began to burst, while others were of much smaller 
size. In the latter case, a longitudinal section of a germ 
showed that it had died in the differentiation phase. It 
looked like a small (1.5-2.0 mm) brown protuberance. 
After the degeneration of the germinal cone the bud dries 
up and abscisses. Abortive female buds leave small (2.5- 
3.0 mm) dark triangular traces (fig. la). 

Pollination in Siberian cedar pine generally occurs in 
the second half of July (Nekrasova 1972). After pollina- 
tion the cones grow for some time until they are 2 mm 
long. During this period (June-July) the cones may also 
cease their development and abort. Traces of cones that 
abscissed shortly after the pollination are larger (4-5 mm), 
and more roundish than those of the abortive female buds 
and are relatively smooth surfaced (fig. lb). 



In the year of maturation both the cones and shoots 
grow larger. The cones reach their ultimate dimensions 
by the time of fertilization (Jvme 15-20) (Nekrasova 1972). 
Therefore, the cones that absciss in the process of matura- 
tion (June-August) leave the traces of the same shape and 
size as normal mature cones do. The traces are large (up 
to 7-8 mm) and oval with a funnel-shaped inner surface 
(fig. Ic, d). Mature cones absciss in late August or early 
September. The inner surface of their traces on the shoot 
bark is covered with resin (fig. Id). 

TRACE IDENTIFICATION 

The identification of cone traces would require knowl- 
edge of the morphostructure of the mature shoot and the 
origin of its different parts. The annual shoot in large 
cone-bearing branches in Siberian cedar pine consists of 
two elementary shoots (Vorobjev and others 1989). The 
first (spring) shoot develops from the winter bud in June. 
The second (summer) shoot is initiated in June and is 
found to enlarge at the beginning of Jvdy. Each elemen- 
tary shoot has short shoots (brachyblasts) and lateral ex- 
tending shoots. The length of the axis and the nvunber of 
its metameres in the spring shoot are approximately ten 
times as large as in the summer shoot. The female cones 
are located in the distal part of the spring shoot. Small 
lateral vegetative shoots usually grow at the same level 



202 



as the female cones. However, there may be no lateral 
shoots. The main whorl of large lateral shoots is located 
on the simimer shoot. It shoiild be noted that simimer 
shoots in mature trees are formed when the spring shoot 
bears cones. Therefore, the summer shoot can be regarded 
as an indirect evidence of the presence of cones in the 
spring shoot. The traces of mature cones on mature need- 
less shoots will occur in the distal part of each annual 
shoot under the main whorl of the lateral shoot. 

The traces should be counted from the top of the branch 
to yield the number of different types of traces on each 
annual shoot. This procediire makes it possible to recon- 
struct the entire dynamic cycle of the initiation, develop- 
ment, and matiiration of cones (over 10 to 12 years). The 
traces of abortive female buds on older shoots are de- 
stroyed because of the secondary bulge of the shoot axis 
and core cracking, which precludes their identification 
and count. Large traces of mature cones persist for a 
longer period of time, and they can be used for the recon- 
struction of crop dynamics over approximately 20 years. 

SAMPLING 

Our experience shows that it is necessary to sample 
10 branches from one tree. The branches selected must 
be from different parts of the crown (top, middle, lower). 
Each branch must be no less than 20 years old. To recon- 
struct the djTiamics of the initiation and development of 
cones on individual branches, one has to use branches of 
mediimi size and ignore the traces of cones on the lateral 
shoots. On the other hand, the seminiference dynamics 
for the tree or stand as a whole can be followed by using 
cone-bearing branches of different size and, most impor- 
tant, by taking into account the traces on small lateral 
branches. 

As the number of annual rings increases, the traces 
of cones on the bark in the cone-bearing branches of trees 
become indiscernible. Therefore, a procedure was devel- 
oped for the identification of such traces on the cross- 
sectional saw cuts of branches (Vorobjev 1979). The saw 
cuts of the bases of whorls show that the traces of cones 
remain within shoots and are quite discernible, which 
makes it possible to reconstruct the crop dynamics virtu- 
ally for the entire Ufe of the tree. Thus, the time scale of 
the proposed procedure can be extended. 

The saw cuts must be made 1 cm beneath the point 
where the longitudinal axis of the branch intersects the 
axes of the lateral shoots. Then they are paper-sanded 
and examined with a stereomicroscope x 16. The saw cuts 
must begin with the whorls where the external traces on 
the bark are fairly visible. The identification of the inner 
traces should also begin with those whorls. This is espe- 
cially important for the differentiation of different periods 
of cone development. 

TRACE CHARACTERISTICS 

The cone traces on the saw cuts of branches can be 
described as follows: The trace of the cone that abscissed 
before pollination (the trace of the abortive female bud) 
has the form of a brown ray with a uniform width of 



0.3±0.03 mm and a length of 1.6±0.2 mm. It runs from 
the pith of the shoot to its bark at the backgrotmd of a 
yellow saw cut wood (fig. 2a). The trace of the cone that 
abscissed after pollination and that of the immature cone 
are half as large edgewise as the trace of the cone that 
abscissed before pollination. The latter trace (fig. 2b) in 
contrast to the former (fig. 2c) has a bulge at the end (1.7 
times as large as the ray width) or is slightly forked. The 
trace of the mature cone (fig. 2d) is twice as large as that 
of the immature cone. Note that there may be no ray. 
The trace of the mature cone differs greatly from all the 
other traces (see fig. 2d and figs. 2a, b, c) by an oval or 
triangular bulge (2.4+0.3 mm wide) at the ray end. Such 
traces are often accompanied by a specific deformation of 
the tree rings as a result of rupture of phloem and xylem 
tissues of the shoot by a great n;unber of cones. Thus the 
more mature the cone the brighter and longer is its trace 
on the saw cut. The bulge is also larger. 

The traces of cones differ from the traces of the lateral 
shoots and brachyblasts on the cross-sectional saw cuts. 
The trace of the lateral shoot (fig. 2e) is wide and gener- 
ally runs toward the bark to join the shoot axis. The trace 
of the cone does not reach the bark. The trace of brachy- 
blast (fig. 2f) is very similar to the trace of the cone that 
abscissed before pollination (the abortive female bud), but 
the former trace is much longer (2.5+0.2 mm) and its end 
grows narrower. The main thing is that the trace of 
brachyblast is surrounded by a brighter layer of wood 
(3.8±0.3 X 0.7±0.1 mm), which is visible to a naked eye. 

USE OF METHODS 

We feel that use can be made of methods based on the 
difference in the density of the cone traces and that of the 
surrounding wood (roentgenography, densitography, laser 
techniques, etc.) in the cases where the traces on the saw 
cuts are dim (especially the traces of the prematvirely 
abscissed cones), or else, if the saw cut has been made in 
a wrong way. In particular, soft x-ray roentgenography 
of a 5- to 7-mm-thick saw cut of a branch (the time of expo- 
sure was 6 minutes) made it possible to determine unam- 
biguously the structure of tree rings, cone traces, and those 
of other metameres (Vorobjev 1979). The above techniques 
provide docvmientary evidence for subsequent analysis. 

This ability to distinguish the internal traces must be 
practiced using the branches with fairly visible external 
cone traces on the bark. The selection of the branches 
for the analysis must be made in the female zone of the 
crown. The branches with one top are most suitable for 
examining the relationship between the seminiference 
and the shoot growth. For the study of crop dynamics, 
on the other hand, the branches with a few tops must be 
used. These branches often occur in the lower part of the 
female zone. 

Sample branches of the tree must be taken from three 
age groups (30, 60, and 100 years old). The extreme val- 
ues of the crop capacity (the number of cones on one 
shoot) are omitted in the case of young branches for the 
first few years before cone bearing and in the case of old 
branches for the last tens of years after cone bearing. The 
extremes are ignored because of the absence of cone 



203 







Figure 2 — Traces of generative and vegetative 
organs on cross-sectional saw cuts of branches. 
X 3: (a) cone that abscissed before pollination, 
(b) cone that abscissed after pollination, (c) cone 
that abscissed during maturation, (d) mature 
cone, (e) lateral shoot, (f) brachyblast. 



traces. The information is taken into account within the 
number series, for it reflects nonseed years. The traces 
are hardly visible at this time because of the predominant 
abscission of 1-year-old conelets caused by early frosts or 
the lack of pollination. 

No less than three to five 200- to 250-year-old trees of 
the first or second Kraft's class are to be selected from one 
sample area for the study of sample branches. 

A retrospective study of the seminiference of Siberian 
cedar pine on the basis of the proposed method makes it 
possible to assess the effect of the reproductive activity of 
trees on the annual ring structure and size, to follow the 
crop cyclicity, to elaborate the ways of the long-term and 
super-long-term forecasting of crops, and to construct an 
ontogenetic model of the relationship between the growth 
and seminiference of shoots. This method appears appli- 
cable to other coniferous plants and, first and foremost, 
to Pinaceae. 

REFERENCES 

Gorchakovsky, P. L. 1958. New issues in the method of in- 
vestigation of the seminiference dynamics of coniferous 
trees. Bot. Zhourn. 43(10): 1445-1459. [In Russian]. 

Iroshnikov, A. I. 1963. The cone-bearing of Siberian cedar 
pine in western Sayani. In: Cone-bearing of Siberian 
cedar pine in eastern Siberia; Moscow: Publ. of USSR 
Academy of Science: 104-119. [In Russian]. 

Korchagin, A. A. 1960. Methods of account of seminiference 
of trees and forest communities. In: Field Geobotany; 



Moscow-Leningrad: Publication of USSR Academy 
of Science: Vol. 2: 41-133. [In Russian]. 

Nekrasova, T. P. 1957. On the methods of investigation 
of cone-bearing dynamics in coniferous trees. Izv. Vost. 
Fil. Akad. Nauk SSSR. 6: 138-145. [In Russian]. 

Nekrasova, T. P. 1972. Biological principles of seminif- 
erence of Siberian cedar pine. Novosibirsk, Nauka: 
272 p. [In Russian]. 

Nesterov, N. S. 1914. On the problem of the methods of 
investigation of cone-bearing of trees. Forest-Industrial 
Vestnik. Vol. 26. [In Russian]. 

Renvall, A. 1912. Die periodischen Erscheinungen der 
Reproduction der Kiefer an der polaren Waldgrenze. 
Acta Forest Fenn. 22(3): 130 p. 

Trofimova, Z. I. 1953. Determination of pine crops by bio- 
logical method. For. Econ. 1: 67-69. [In Russi£Ui]. 

Vorobjev, V. N. 1964. The cone-bearing of Siberian cedar 
pine in altitude subzones of north-eastern Altai. Izv. 
Sib. Otd. Akad. Nauk SSSR, Ser. Biol.-Med. Nauk. 
12(3): 86-90. [In Russian]. 

Vorobjev, V. N. 1974. Peculiarities of cone-bearing of 
Siberian cedar pine in mountain conditions. In: Biology 
of seed reproduction of coniferous trees of western 
Siberia. Novosibirsk, Nauka: 15-70. [In Russian]. 

Vorobjev, V. N. 1979. A method of retrospective study 
of seminiference dynamics in Pinus sibirica Du Tour 
(Pinaceae). Bot. Zhourn. 64(7): 971-974. [In Russian]. 

Vorobjev, V. N.; Vorobjeva, N. A.; Goroshkevich, S. N. 
1989. Growth and sex of Siberian cedar pine. 
Novosibirsk, Nauka. 167 p. [In Russian]. 



204 



1 ^ \ \ 

Importance to Wildlife 




International Workshop 
St. Moritz 1 992 

^ — ^ 



205 



NUTRITIVE VALUE OF WHITEBARK 
PINE SEEDS, AND THE QUESTION OF 
THEIR VARIABLE DORMANCY 

Ronald M. Lanner 
Barrie K. Gilbert 



Abstract— Seeds of whitebark pine (Pinus albicaulis) are eaten 
year-round by their disperser, Clark's nutcracker (Nucifraga 
Columbiana), and may be the only food that is fed to their young. 
They are also preferentially larder-hoarded by the red squirrel 
{Tamiasciurus hudsonicus); and are eaten in large quantities by 
black bears (Ursus americanus) and grizzly bears {U. arctos), es- 
pecially prior to hibernation. 

Analyses show whitebark pine seeds have a mean seed weight 
of 119 mg, of which kernels comprise 52.5 percent. Kernel content 
was approximately 21 percent carbohydrate, 21 percent protein, 
52 percent fat, 3 percent ash, and 3 percent water. Major mineral 
nutrients present were Cu, Zn, Fe, Mn, Mg, and Ca. Sixteen amino 
acids were present, with glutamic acid, lysine, and arginine espe- 
cially abundant. Major fatty acids were linoleic, oleic, and y-linolenic, 
with lesser amounts of 11 others. The unsaturated fraction was 
92 percent. 

The variable period of dormancy is analyzed in terms of its con- 
tribution to fitness in a variable environment, its phylogenetic 
distribution within Pinus subg. Strobus and subsect. Cembrae, 
and its effect in retaining nutcracker populations by broadening 
the food base. The food base is broadened by the interaction of 
memory duration, seed crop frequency, and foraging on buried 
dormant seeds whose location is betrayed by presence of germi- 
nants. This may explain the excessive caching behavior of nut- 
crackers and may reduce the frequency of irruptions during times 
of food shortage. Even when seed crops are sporadic, variable dor- 
mancy can make seeds available to nutcrackers in most years, and 
as early as the first year after dispersal birds can forage on seeds 
cached by other birds. 



Whitebark pine (Pinus albicaulis Engelm.) is the only 
North American member of Pinus subsect. Cembrae, the 
stone pines. The species is found in subalpine habitats in 
the Northern Rocky Mountains of Canada and the United 
States, and in the Cascade Mountains-Sierra Nevada sys- 
tem (Critchfield and Little 1966). Because whitebark pine 
usually grows in inaccessible areas and has relatively low 
commercial value, it was long neglected by researchers. 
But in recent years, due largely to interest in corvid-pine 
mutualisms stimulated by a seminal review (Turcek and 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

Ronald M. Lanner is Professor, Department of Forest Resources, and 
Barrie K. Gilbert is Senior Scientist, Department of Fisheries and Wildlife, 
both at Utah State University, Logan, UT. 



Kelso 1968), it has received greatly increased attention. 
Much of this attention has focused on the important role 
of whitebark pine seed in the biology of animals that in- 
habit the whitebark pine forest. 

It is the large Upid-rich "nut" of this pine that attracts its 
seed disperser, Clark's nutcracker (Nucifraga columbiana) 
(Tomback 1978). Evidence indicates that the pine can reU- 
ably regenerate only from seeds cached in the soil by its nut- 
cracker mutualist (Hutchins and Lanner 1982), and that 
these seeds are the sole nutritional resource for nutcrackers 
throughout their juvenile stage (Vander Wall and Hutchins 
1983). Further, it has been shown that whitebark pine 
seeds are a preferred food of the red squirrel (Tamiasciurus 
hudsonicus), which larder-hoards great quantities of cones 
on and below the groimd (Hutchins and Lanner 1982). Both 
black bears (Ursus americanus) and grizzly bears (U. arctos) 
raid the squirrel middens in order to gorge themselves with 
pine nuts both before and after hibernation (Kendall 1983; 
Mattson and Jonkel 1990). Thus the whitebark pine seed 
is of obviously great nutritional importance to at least five 
organisms of its ecosystem: Clark's nutcracker, the red 
squirrel, black and grizzly bears, and whitebark pine itself, 
for whose benefit the seed originally evolved. It is therefore 
of more than passing interest to know the nutrient content 
of this valued seed. 

Pubhshed analyses of whitebark pine nuts have stressed 
their high caloric value, which is mainly a function of high 
lipid content. Emphasis on caloric content appefirs to be a 
response to the current ecological fixation on energetics of 
food resources, almost to the exclusion of other food charac- 
teristics. Thus Lanner (1982) has documented the parti- 
tioning of energy in nuts and other cone components, and 
Hutchins and Lanner (1982) have reported changes in the 
caloric value of whitebark pine seeds as the seeds mature, 
but other aspects of the nutritional qualities of these seeds 
have been neglected. 

Another characteristic of the whitebark pine seed that 
requires attention is its variable dormancy. Seeds of several 
Cembrae species have been reported to delay germination 
until the second or third year following dispersal (USDA 
FS 1974). This has been attributed to the immaturity of 
embryos at the normal germination date, and has been 
blamed for low germination rates of whitebark pine seeds 
in laboratory tests (Pitel and Wang 1990). Kozlowski and 
Gunn (1972) have suggested that germination delays ensure 
estabhshment "even though the early germinants failed to 
survive severe environmental stresses such as droughts or 
severe frosts," but they offer no data to support this intu- 
itively attractive idea. According to McCaughey (1992), 
whitebark pine seeds of a cohort displayed germination 



206 



dming the first (11 percent), second (45 percent), third 
(11 percent), and fourth (percent not specified) postdis- 
persal years. 

The objectives of this paper are to document the nutri- 
tional characteristics of whitebark pine seeds, and to sug- 
gest some possible ramifications of the variable dormancy 
habit. 

SEED COMPOSITION ANALYSES 

Analyses were made of whitebark pine seeds of the 1989 
crop fi-om around Gardiner, MT (2,635 m elevation), provided 
by Dr. Ward McCaughey of the Intermountain Research Sta- 
tion (USDA Forest Service), Bozeman, MT. 

Protein determination was made by the Kjeldahl method, 
fat determination by the chloroform methanol method of 
Hubbard (1977), and mineral assays by atomic absorption 
by Drs. Arthiir Mahoney and Dejia Zhang, Department of 
Nutrition and Food Sciences, Utah State University. Five 
samples were used for these determinations. Amino acid 
analyses of nut protein were made by Dr. Rod Brown, De- 
partment of Nutrition and Food Sciences, Utah State Uni- 
versity, using a Beckmann 6300 amino acid analyzer. Four 
nms were made: data presented here are the means of those 
runs. Fatty acid composition of kernel fats was provided by 
Laurence G. Cool of Dr. E. Zavarin's laboratory at the For- 
est Products Laboratory, University of Cahfomia, Richmond, 
CA, and was determined by gas chromatography. 

Results 

Kernel Composition — Mean values of 20-seed samples 
yielded the following seed coat versus kernel (female game- 
toph5^e plus embryo) breakdown: seed coat, 47.5 percent, 
0.0568 g; kernel, 52.5 percent, 0.0626 g. Kernels consisted 
of water (3.5 percent), ash (2.8 percent), carbohydrates 
(21 percent) [calculated value], protein (21 percent), and 



Table 1 — Amino acids identified in kernel 
protein of whitebark pine in order 
of abundance (mean of four runs) 



Amino 


Nanomoles 


acid 


per mg 


Glutamic acid 


0.729 


Lysine 


.666 


Arginine 


.554 


Aspartic acid 


.422 


Glycine 


.386 


Valine 


.380 


Alanine 


.373 


Leucine 


.369 


Serine 


.363 


Proline 


.352 


Isoleucine 


.245 


Phenylalanine 


.192 


Tyrosine 


.168 


Threonine 


.147 


Histidine 


.126 


Methionine 


.038 



Table 2 — Comparative abundance of amino acids identified in 
kernel protein of three pine species^ 



Amino 
acid' 


P. albicaulis 


P. edulis 


P. pinea 


^1 1 1 


1 An 


0 / 


ft7 


1 vo 
LYb 


y 1 


1 Q 




A 

Ano 


fO 


1 nn 

1 UU 


1 nn 


ACD 


RQ 
OO 




A7 


vaLY 


til 




OR 


VAL 




07 




Al A 
MLA 


Dl 


01 


97 


1 CI 1 


01 


O / 


Ol 


CCD 
OtPl 


OU 


00 


OO 
r r 


rnU 


AO 


OQ. 




loU 


'>A 

o4 


OCi 


on 


Due 

rn t 




on 


on 


1 Yn 


23 


H n 


i» 


THR 


20 


16 


16 


HIS 


17 


12 


12 


MET 


5 


9 


9 


CYS 




9 


9 


TRY 




6 


1 



'Species and references: P. albicaulis (this paper), P. edulis (McCarthy and 
Matthews 1984), P. pinea (McCarthy and Matthews 1984). 

^Amino acids not previously mentioned; CYS = cystine, TRY = tryptophan. 



fat (52 percent). Atomic absorption analysis disclosed the 
following mineral composition, in micrograms per gram of 
ash: Cu, 17.5; Zn, 66.9; Fe, 35.7; Mn, 132; Mg, 17.5; Ca, 51.9. 

Amino Acids — Sixteen amino acids were identified in the 
whitebark pine kernel samples. The most abimdant were 
glutamic acid, lysine, arginine, and aspartic acid (table 1). 
These ranged from 0.42 to 0.73 nmol/mg. Also abxmdant, 
ranging from 0.35 to 0.39 nmol/mg were glycine, valine, ala- 
nine, leucine, serine, and proline. Lesser quantities were 
found of isoleucine, phenylalanine, tyrosine, threonine, and 
histidine; and least of methionine (table 1). 

Amino acid quantities have been expressed in the Htera- 
ture in various ways. Therefore, to compare the amino acids 
of whitebark pine with those of other species, we have given 
the most abimdant amino acid of each species the index 
value of 100, and aU others are scaled appropriately within 
each species. Thus, while the most abimdant amino acid 
(100) of whitebark pine is glutamic acid, that oiPinus edulis 
and P. pinea is arginine (table 2). 

Fatty Acids — Fifteen fatty acids were identified on the 
chromatograph. The most abimdant of these were the poly- 
unsaturated linoleic acid and monounsaturated oleic acid, 
which comprised 46.9 percent and 30.6 percent of the total 
fatty acid methyl esters, respectively. The other components 
are present in much smaller proportions (table 3). The total 
fraction of imsaturated fatty acids was 92.0 percent, 32.1 
percent monoimsaturated, and 59.9 percent polyunsaturated. 

Discussion 

Due to differences in analytical techniques, and variation 
in the multipHer used in determining protein content, the 
pubUshed fractions of fats, proteins, carbohydrates, and 
ash in nut kernels are quite variable. 



207 



Table 3 — Fatty acids identified in kernel fat of 
whitebark pine in order of abundance 





Total fatty acid 


Fatty acid 


methyl esters 




Percent 


Linoleic 


46.9 


Oleic 


30.6 


y-Linolenic 


9.9 


Palmitic 


4.2 


Stearic 


2.7 


1 1 -Eicosenoic 


1.4 


y-Linoleic 


1.2 


5-, 11-, 14-Eicosatrienoic 


.8 


1 1 -, 1 4-Eicosadienoic 


.7 


Arachidic 


.6 


Linolenic 


.3 


DOl lc?l IIU 


o 


Vaccenic 


.1 


5-, 1 1 -Eicosadienoic 


.1 


Unknowns 


.3 


Total 


100.0 



Whitebark pine's protein content of 21 percent compares 
with pinyon pine values in the range of 10 percent (Pinus 
monophylla) to 19 percent (P. cembroides) (Lanner 1981); 
25 percent in P. sabiniana, and 31 percent in P. pinea 
(Farris 1983). 

The proteins in pine nuts are almost entirely seed storage 
proteins of the megagametophyte. Following the hydrolysis 
of these proteins, the component amino acids are exported 
into the growing points of the embryo (Lammer and GifiFord 
1989). GifiFord (1988) has shown there is considerable varia- 
tion in electrophoretic characteristics of seed proteins of 
several species (including P. albicaulis); and Schirone and 
others (1991) have shown such variation to have taxonomic 
significance. It is therefore surprising to see the almost- 
identical gmaino acid profiles of P. pinea and P. edulis seed 
proteins (table 2). These species are in separate subgenera 
of Pinus (Critchfield and Little 1966) and would be expected 
to have proteins of quite different composition. 

Foiu" common amino acids were missing fi-om whitebark 
pine seed protein. Hydroxyproline's absence can probably 
be attributed to a lack of that amino acid in embryo and 
storage protein, and its restriction to seed coats and peri- 
carp material (Van Etten and others 1967). Tr3T)tophan is 
generally a minor component of plant proteins (White and 
others 1959) and can be destroyed in analysis (Farris 1983). 
Neither cystine nor cysteine were identified in whitebark 
pine, though the former is found in small quantities in 
P. edulis and P. pinea (table 2). The lack of cystine and the 
low quantity of methionine indicate that sulfur-containing 
amino acids are of little importance to development of white- 
bark pine seedlings. 

Perhaps the most striking difference between seed protein 
of whitebark pine and the other species in table 2 is its great 
abundance of lysine. Lysine, an essential amino acid for 
human growth, is notoriously low in cultivated members of 
the Graminae. Feeding experiments conducted with dark- 
eyed juncos iJunco hyemalis) by Parrish and Martin (1977) 



have established the essentiality of lysine in maintaining 
a positive nitrogen balance in that bird. Additional amino 
acids found essential for poultry are arginine, glycine, £ind 
histidine, all of which are found in whitebark pine nuts 
(table 1). In the absence of data specific to the genus Nuci- 
fraga, we speculate that these nuts contain all the amino 
acids essential to nutcrackers. 

According to Van Etten and others (1967) the nutrition- 
ally essential amino acid requirements for swine are similar 
to those of humans. If the requirements for bears are also 
reasonably similar, then whitebark pine protein, as well as 
its fat, may be of great dietary significance for bears. Bears 
are more likely to satisfy their protein needs firom a wider 
variety of foodstuffs than nutcrackers, however. 

The fat content of whitebark pine nuts in this study (52 
percent) is qmte similar to that of other stone pines as deter- 
mined by Tikhomirov (1939): P. cembra, 59.9 percent and 
P. pumila, 59.4 percent. Fat values for P. sibirica vary over 
the range 51 to 75 percent (Shimanjoik 1963). The fat con- 
tent of pinyon pines (Lsmner 1981) ranges fi'om 23 percent 
in P. monophylla to 62 to 71 percent in P. edulis. Farris 
(1983) reports that P. sabiniana kernels contain 49 percent 
of fat, and P. pinea contains 47 percent. 

Tikhomirov (1939) reported the fat content of P. koraien- 
sis to be 52.5 percent, but recent work by Yoon and others 
(1989) reports a crude lipid fi'action of 72.5 percent, which 
reduced to 70.4 percent when purified. 

The most notable features of the whitebark pine seed's 
fatty acids are the preponderance of oleic (30.6 percent) and 
linoleic (46.9 percent) acids. These are both unsaturated 
fatty acids, and it appears they are commonly the major 
ingredients of the seed fatty acids (f.a.) of Pinus. Thus they 
comprise 28.4 percent and 47.9 percent of the f.a. of P. kor- 
aiensis, 40.0 percent and 41.9 percent of the f.a. of P. edulis 
(R. M. Lanner, impubl. data), and 42.5 percent and 42.0 per- 
cent of the f.a. of P. monophylla, respectively (R. M. Lanner, 
impubl. data). According to Yoon and others (1989) the f.a. 
of P. koraiensis are 8.75 percent saturated, 29.41 percent 
monounsaturated, and 61.64 percent poljomsaturated. Pinus 
edulis and P. monophylla have over 80 percent polyunsatu- 
rates. The nut of P. sabiniana was found by Farris (1983) 
to be 50.5 percent oleic and 45.2 percent linoleic acids — 
over 95 percent polyimsaturated. 

In addition to containing very high ratios of imsaturated 
to saturated fatty acids, pine nuts are also fi'ee of cholesterol 
(McCarthy and Matthews 1984). Yoon and others (1989) 
point out that P. koraiensis nuts have long been used in 
Korean medicine, and the nut oil of P. cembra had medici- 
nal uses in Poland during the 17th century (Bialobok 1975). 
Yoon and others point out that P. koraiensis seeds contain 
omega-5 fatty acids, the functions of which in the himian 
body are unknown. They suggest that these f.a. may be 
beneficial to health, and should be studied in relation to 
lipid and glucose metabolism. 

VARIABLE DORMANCY 

The variable dormancy that characterizes whitebark 
pine seeds has been viewed as a bothersome habit that 
makes the nurseryman's task more difiScult. However, it 
could well be a trait that has an important biological func- 
tion and that has been naturally selected to perform that 



208 



Table 4^Number of years during a 15-year period in which seeds will be germinating, under assumptions of different seed-crop intervals, and 
of dormancy varying from 0 to 3 years^ 



Period of seed 

dormancy Interval between successive seed crops 





1 


2 


3 


4 


5 


6 


7 


8 


9 


10 


11 


12 


13 


14 


15 


0 


15 


8 


5 


4 


3 


3 


3 


2 


2 


2 


2 


2 


2 


2 


1 


1 


15 


15 


10 


8 


6 


6 


5 


4 


4 


4 


4 


4 


4 


3 


2 


2 


15 


15 


15 


12 


9 


9 


7 


6 


6 


6 


6 


6 


5 


4 


3 


3 


15 


15 


15 


15 


12 


11 


9 


8 


8 


8 


8 


7 


6 


5 


4 



'Seeds dormant for 0 years all germinate the spring following dispersal. Seeds domiant more than 0 years germinate in part in the spring following disperal; and 
in part during each additional spring for 1 , 2, or 3 years. 



function. Here we postulate and examine several possible 
evolutionary functions for variable dormancy. 

Increased Fitness 

According to this hypothesis, the conditions met by ger- 
minating seeds may vary greatly from year to year. There- 
fore a cohort of seeds dispersed in fall of the year n may be 
confronted by disadvantageous — or even catastrophic — 
conditions in spring of n+1. Such conditions could result 
from severe frost or drought (Kozlowski and Gunn 1972). 
If so, it might be to the plant's advantage to delay germi- 
nation of some seeds until spring of n+2 or even n+3 or be- 
yond. If the temporal spacing of seed crops and the lengths 
of the variable dormancy periods are matched properly, par- 
tial seed cohorts could be germinating nearly every year. 
Under such a sceneirio there would be a high probability of 
some seeds germinating during the years of most favorable 
conditions. 

As an example, assimie a pine stand bearing seeds in the 
first year of an arbitrarily chosen 15-year period. If seed 
crops reoccur every 4 years (in years 5, 9, and 13 of this 
15-year period), and if germination occiu-s only in the spring 
foUovdng dispersal (spring of years 2, 6, 10, and 14), then 
there will be 4 years during the 15-year period when seed- 
lings can become established (table 4). However, if some 
of those seeds remain dormant for variable lengths of time, 
germinating not only in the spring following dispersal, but 
in the next three springs as well, seedlings can become es- 
tablished during all 15 years of the period (table 4). Even 
if seed crops were spaced at 10-year intervals, this level of 
variable dormancy would produce seedlings in 8 of the 15 
years. Thus it is apparent that variable dormancy can neu- 
tralize the effects of infrequent crop years by creating a 
steady stream of germinating seeds during and between 
crop years. 

If variable dormancy increases whitebark pine's fitness by 
increasing the frequency of germination years, one might ex- 
pect associated conifers — especially other pines — ^to show the 
same dormancy behavior. Apparently, however, this seems 
not to be the case. If we assume that stratification treat- 
ments prescribed to overcome seed dormancy are an index 
of a seed's behavior in nature, it is apparent that the subal- 
pine conifers of western North America show no consistency 
(table 5). This does not mean that whitebark pine derives 
no benefits from being able to provide germinants in most 



years, but perhaps that only whitebark pine (and maybe 
P. monticola) enjoys those benefits. Apparently variable 
dormancy is not a requirement for success of whitebark 
pine's associates in the subsdpine zone. In fact, variable 
dormancy might be detrimental to these other conifers, 
because, except for P. flexilis, they have wind-dispersed 
seeds that lie on the soil surface, where they are vulner- 
able to predation. Buried seeds of whitebark pine (and lim- 
ber pine) would be more likely to survive an additional year 
or more of "waiting" to germinate. 

If variable dormancy is a genetically controlled charac- 
ter, it would be important to learn its mode of inheritance. 
For example, do all the seeds of a tree exhibit the same 
dormancy behavior, as would be expected in a maternally 
transmitted characteristic of seed-coat or megagametophytic 
tissue; or do they segregate, as would be expected if inher- 
itance from both parents is manifested in variable embryo 
behavior? If the latter, is genetic control due to major gene 
effects, or is control achieved multigenically? Seed tests 
would need to be made with the products of controlled 
crosses to unravel this and similar questions, and careful 
anatomical studies would have to be made to better imder- 
stand embryo growth and development. 



Table 5 — Recommended^ cold stratification periods for seeds of 
subalpine conifers of western North America and their 
overlap with that of Pinus albicaulis 

Overlap of 

Stratification P. albicaulis 

Species period stratification period 

Days Percent 



Pinus albicaulis 


90-120 




P. monticola 


30-120 


100 


P. flexilis 


21 - 90 


3 


P. aristata 


0 


0 


P. balfouriana 


90 


3 


P. contorta var. latifolia 


30- 56 


0 


P. contorta var. murrayana 


20- 30 


0 


Tsuga mertensiana 


90 


3 


Picea engelmannii 


0 


0 


Abies lasiocarpa 


28 


0 


A. amabilis 


21 - 28 


0 


A. magnifica 


28- 42 


0 



In USDA FS1974. 



209 



Phylogenetic Legacy 

Perhaps whitebark pine exhibits variable dormancy sim- 
ply because the trait is conmion to a group of related species 
that includes whitebark pine. To determine this, we have 
displayed in table 6 the available stratification requirements 
of pines of Pinus subg. Strobus by subsection. 

Species characterized (in USDA FS 1974) as "suspected 
of having immature embryos at the time of collection" or of 
exhibiting "extreme dormancy" are identified by superscript 
nvunbers. The densest concentration of superscripts is in 
the stone pines — subsection Cembrae — in which all species 
require more than 60 days of stratification to germinate 
(table 6). Thus it seems likely that variable dormancy is 
not just an albicaulis trait, nor a broad Strobus trait, but 
a subsection Cembrae trait. This is consistent with my ear- 
lier suggestion (Lanner 1980) that the Cembrae pines are 
of monophyletic origin. 

Nutcracker Retention 

The importance of variable dormancy may lie in its ef- 
fects on whitebark pine's seed disperser, Clark's nutcracker. 
Whitebark pine depends on the nutcracker for seed dispersal 
and seedling establishment (Hutchins and Lanner 1982), 
so it is in the long-term interest of the pine to retain nut- 
cracker populations in its ecosystem. We suggest that vari- 
able dormancy can have that effect. 

The following scenario illustrates our argument. A mast 
year occurs in whitebark pine (year n). The local nutcracker 
population frantically harvests and caches several times 
as many seeds as are required to ensure its own survival 
and that of its offspring imtil the next fall (Hutchins and 
Lanner 1982; Tomback 1978; Vander Wall and Balda 1977). 
Throu^out the fall, the nutcrackers recache their buried 
seeds (Hutchins and Lanner 1982). The next summer, in 
year n+1, some seeds germinate. The birds now have three 
sources of food — their own caches, fi'om which they continue 
to feed; the seeds now germinating and therefore visible; 
and the still-dormant seeds fi'om the same caches, whose 
locations are betrayed by the n+1 germinants (Vander Wall 
and Hutchins 1983). Each bird now knows not only the 
locations of all its own caches, but also the caches made 
by other birds in which n+1 germinants were found. Even 
if there is no crop in the second year, the bird's 9-month 
memory (Balda 1992) of cache location will allow it to sur- 
vive until the third year, when a new crop may be present. 
Again, sprouting seeds of n+2 will betray the locations of 
remaining caches. If there is again no crop, or only a light 
one, the nutcrackers may be able to survive by recaching 
the dormant seeds they have found associated with germi- 
nating seeds. This assumes that another 9-month period 
of cache remembrance follows each caching event. Due to 
the overlapping of periods of availability of current-crop 
seeds and seeds of the previous crop still lying dormant up 
to 3 years in caches, plus those germinating, nutcrackers 
should be able to remain in situ except in years following 
several successive crop failures. 

Such a scenario could not occur with seeds that all 
sprouted in n+1. With variable dormancy it appears to be 
a strong possibility; however, intensive observation and 
additional data will be required to determine whether it 



Table 6— Recommended^ cold stratification periods for seed of 

some pines in subgenus Strobus, and their overlap with 
that of Pinus albicaulis 



Species 



Stratification 
period 



Overlap of 
P. albicaulis 
stratification period 



Days Percent 

(Pinus) 

albicaulis^ 90-120 — 

subsection Strobi 

flexilis^ 21-90 3 

monticola^ 30-120 100 

strobus 60 0 

strobiformis^ 60-120 100 

lambertiana^ 60- 90 3 

peuce 0-60 0 

parviflora^-^ 90 3 

griff ithii 0-15 0 

subsection Balfourianae 

aristata 0 0 

balfouriana^ 90 3 

subsection Cembroides 

cembroides 0 0 

edulis 0 0 

monophylla^ 28- 90 3 

quadrifolia 0 0 

subsection Cembrae 

cembra^'^ 90-270 100 

s/b/r/ca" 60- 90 3 

pumila^ 120-150 3 

koraiensis^'^ 90 3 

'In USDA FS 1974. 

^Characterized as exhibiting "extreme dormancy" (requiring more than 60 
days of stratification in USDA FS 1974). 

'Characterized as "suspected of having immature embryos at the time of 
collection" in USDA FS 1974. 



is realistic. Our intention in speculating is to stimulate 
fiirther study of this possibility. All of this illustrates that 
to understand whitebark pine's variable seed dormancy, 
the nutcracker too must be considered. As pointed out by 
Lanner (1980): "The ecology and evolution of the wingless- 
seeded 'Bird Pines' can only be understood in a frame of 
reference that includes the behavior of their feathered dis- 
persers." The bare facts of the mutualism of whitebark 
pine and Clark's nutcracker are by now well established. 
The time is ripe to study the complexities of the relation- 
ship, and its second-order effects. 

REFERENCES 

Balda, R. P. 1992. [Conversation with Ronald M. Lanner]. 
August 14. Flagstaff, AZ: Northern Arizona University. 

Bialobok, S., ed. 1975. Stone-pine (Pinus cembra L.). Trans- 
lated from Polish for U.S. Department of Agriculture and 
National Science Foundation. National Technicsd Infor- 
mation Service TT-73-54045. 173 p. 

Critchfield, W. B.; Little, E. L., Jr. 1966. Geographic dis- 
tribution of the pines of the world. Misc. Publ. 991. 



210 



Washington, DC: U.S. Department of Agriciilture, Forest 
Service. 97 p. 

Farris, G. J. 1983. California pignolia: seeds of Pinus sabini- 
ana. Economic Botany. 37: 201-206. 

Gifford, D. J. 1988. An electrophoretic analysis of the seed 
proteins from Pinus monticola and eight other species of 
pine. Canadian Journal of Botany. 66: 1808-1812. 

Hubbard, W. D. 1977. Comparison of various methods for 
the extraction of total lipids, fatty acids, cholesterol, and 
other sterols from food products. Journal of the American 
Oil Chemists' Society. 54: 81-83. 

Hutchins, H. E.; Lanner, R. M. 1982. The central role of 
Clark's nutcracker in the dispersal and estabhshment 
of whitebark pine. Oecologia. 55: 192-201. 

Kendall, K 1983. Use of pine nuts by black and grizzly bears 
in the Yellowstone area. In: Meslaw, Charles E., ed. Bears — 
their biology and management. West Glacier, MT: Inter- 
national Association for Bear Research and Management: 
166-173. 

Kozlowski, T. T.; Gunn, C. R. 1972. Importance and charac- 
teristics of seeds. In: Kozlowski, T. T., ed. Seed biology. 
New York: Academic Press: 1-20. 

Lammer, D. L.; Gifford, D. J. 1989. Lodgepole pine seed 
germination. II. The seed proteins and their mobihzation 
in the megagametophyte and embryonic axis. Csmadian 
Journal of Botany. 67: 2544-2551. 

Lanner, R. M. 1980. Avisin seed dispersal as a factor in the 
ecology and evolution of limber and whitebark pines. 
In: Dancik, B. P.; Higginbotham, K O., eds. Sixth North 
American forest biology workshop proceedings; 1980 
August 11-13; Edmonton, AB. Edmonton, AB: University 
of Alberta: 15-45. 

Lanner, R. M. 1981. The pinon pine: a natural and cultural 
history. Reno, NV: University of Nevada Press. 208 p. 

Lanner, R. M. 1982. Adaptations of whitebark pine for seed 
dispersal by Clark's nutcracker. Canadian Journal of 
Forest Research. 12: 391-402. 

Mattson, D. J.; Jonkel, C. 1990. Stone pines and bears. In: 
Schmidt, W. C; McDonald, K. J., comps. Proceedings — 
sjnnposium on whitebark pine ecosystems: ecology of a 
high-mountain resoiu-ce; 1989 March 29-31; Bozeman, 
MT. Gen. Tech. Rep. INT-270. Ogden, UT: U.S. Depart- 
ment of Agriculture, Forest Service, Intermoimtain Re- 
search Station: 223-236. 

McCarthy, M. A.; Matthews, R. H. 1984. Composition 
of foods: nut and seed products. Agric. Handb. 8-12. 
Washington, DC: U.S. Department of Agriculture, Hu- 
man Nutrition Information Service. 137 p. 

McCaughey, W. 1992. [Letter to Ronald M. Lanner]. 
February 25. 1 leaf. Bozeman, MT: U.S. Department 



of Agriculture, Forest Service, Intermountain Research 
Station. 

Parrish, J. W., Jr.; Martin, E. W. 1977. The effect of dietary 
lysine level on the energy and nitrogen balance of the 
dark-eyed junco. Condor. 79: 24-30. 

Pitel, J. A.; Wang, B. S. P. 1990. Physical and chemical 
treatments to improve germination of whitebark pine 
seeds. In: Schmidt, W. C; McDonald, K J., comps. 
Proceedings — sjmiposiiun on whitebark pine ecosystems: 
ecology of a high-moxmtain resource; 1989 March 29-31; 
Bozeman, MT. Gen. Tech. Rep. INT-270. Ogden, UT: 
U.S. Department of Agriculture, Forest Service, Inter- 
mountain Research Station: 130-133. 

Schirone, B.; Piovesan, B.; Bellarosa, R.; Pelosi, C. 1991. A 
teixonomic analysis of seed proteins in Pinus spp. (Pina- 
ceae). Plant Systematics and Evolution. 178: 43-53. 

Shimanyuk, A. P., ed. 1963. Fruiting of the Siberian stone 
pine in east Siberia. Translated from Russian for U.S. 
Department of Agricultural and National Science Foim- 
dation by Israel Program for Scientific Translations. 
Jerusalem. 204 p. 

Tikhomirov, B. A. 1939. Khozyaistvenoye ispol'zovaniye 
kedrovogo stlannika. Sov. Sever. 4: 62-65. 

Tomback, D. F. 1978. Foraging strategies of Clark's nut- 
cracker. Living Bird. 16: 123-161. 

Turcek, F. J.; Kelso, L. 1968. Ecological aspects of food 
transportation and storage in the Corvidae. Communica- 
tions in Behavioral Biology Part A. 1: 277-297. 

U.S. Department of Agriculture, Forest Service. 1974. Seeds 
of woody plants in the United States. Agric. Handb. 450. 
Washington, DC: U.S. Department of Agricultiire, Forest 
Service. 883 p. 

Vander Wall, S. B.; Balda, R. P. 1977. Coadaptations of the 
Clark's nutcracker and the pinon pine for efficient seed 
harvest and dispersal. Ecological Monographs. 47: 89-111. 

Vander Wall, S. B.; Hutchins, H. E. 1983. Dependence of 
Clark's nutcracker, Nucifraga columbiana, on conifer 
seeds dming the postfledging period. Canadian Field 
Naturahst. 97: 208-214. 

Van Etten, C. H.; Kwolek, W. F.; Peters, J. E.; Barclay, A. S. 
1967. Plant seeds as protein sources for food or feed. Eval- 
uation based on amino acid composition of 379 species. 
Journal of Agricultural and Food Chemistry. 15: 1077-1089. 

White, A.; Handler, P.; Smith, E. L.; Stetten, D. 1959. Prin- 
ciples of biochemistry. New York: McGraw-Hill Book Co. 
1149 p. 

Yoon, T.-H.; Im, K-J.; Koh, E. T.; Ju, J.-S. 1989. Fatty acid 
compositions of Pinus koraiensis seed. Nutrition Re- 
search. 9: 357-361. 



211 



BEAR USE OF WHITEBARK PINE 
SEEDS IN NORTH AMERICA 



David J. Mattson 
Daniel P. Reinhart 



Abstract — Whitebark pine (Pinus albicaulis) seeds are an im- 
portant high-quality food for bear populations that occupy ecosys- 
tems with continental climates south of the United States- 
Canada border. Availability of pine seeds affects levels of 
human-bear conflict and bear mortality. In most areas bears 
acquire whitebark pine seeds by excavating red squirrel (Tamia- 
sciurus hudsonicus) food caches. Squirrel densities appear to 
limit bear use of pine seeds more than abundance of whitebark 
pine. Thus management of whitebark pine habitats for bear use 
primarily involves management to favor red squirrels. Bear use 
of pine seeds is restricted to stands >100 years old, and can per- 
sist for an additional 200 to 300 years. Consequently stand rota- 
tions of 300 years and harvest rates of 3 percent per decade are 
recommended. Also, because of potential or demonstrated nega- 
tive impacts associated with global climate change and white 
pine blister rust {Cronartium ribicola), long-term prospects for 
whitebark pine should be part of considerations when managing 
habitats of bear populations, especially insular populations. 



Whitebark pine seeds are a predictably high-quality 
bear food given their relatively leirge size, high fat con- 
tent, and moderately high digestibility (table 1). Fat con- 
tent is perhaps the most important attribute of these 
seeds, given the importance of body fat to the life strategy 
of bears (Stirling and Derocher 1990) and the greater effi- 
ciencies of body fat accimiulation associated with high di- 
etary fat content (Hadley 1985; Schemmel 1976). On the 
other hand, use of whitebark pine seeds especially by griz- 
zly bears (Ursus arctos horribilis) is complicated by the in- 
dehiscent nature of whitebark pine cones (Lanner 1990), 
and the consequent reliance of most bears on vdndthrow 
or rodent intermediaries for access to the seeds (Kendell 
1983; Mattson and Jonkel 1990; Taylor 1964). 

Black (Ursus americanus) and grizzly bears have prob- 
ably made use of pine seeds for as long as these two spe- 
cies have coexisted with whitebark pine, approximately 
2 million years for black bears and 12,000 years for griz- 
zly bears (Axelrod 1986; Km-ten and Anderson 1980). Al- 
though aboriginal Americans were probably aware of this 
relationship, the first written records of bears using 
whitebark pine seeds did not appear until well into the 



Paper presented at the International Workshop on Subalpine Stone 
Pines and Their Environment: The Status of Our Knowledge, St. Moritz, 
Switzerland, September 5-11, 1992. 

David J. Mattson is Wildlife Biologist, Interagency Grizzly Bear Study 
Team, Forestry Sciences Laborestry, Montana State University, Bozeman, 
MT 59717; Daniel P. Reinhart is Management Biologist, National Park 
Service, Resource Management, Yellowstone National Park, WY 82190. 



20th century, in articles by authors such as Cahalane 
(1947), Taylor (1964), and Tisch (1961). Details of the re- 
lationship between bears suid whitebark pine seeds have 
only recently been published, focusing on interior areas 
characterized by more continental climates (Craighead 
and others 1982; Kendall 1983; Mattson and Jonkel 1990; 
Mattson and others 1992, 1993). In this paper we update 
information presented by Mattson and Jonkel (1990), and 
further elaborate on some aspects of the relationship be- 
tween bears and whitebark pine in North America. 

PATTERNS OF BEAR USE 
Geographic Distribution 

Bears consistently make use of whitebark pine seeds 
only south of the United States-Ceinada border, where 
whitebark pine is a potentially abundant component of 
subalpine forests (fig. 1). However, even in this eirea ma- 
jor differences in intensity of bear use are evident, associ- 
ated with maritime and continental climates. 

We examined published bear food habits studies fi-om 
western North America that included the results of 13,130 
bear feces examinations. Our analysis clearly showed 
that the consistency and relative frequency of whitebark 
pine seed use was much greater (18.7 percent) in areas 



Table 1— Size, nutrient content, and relative digestibility by ursids 
of whitebark pine seeds; compiled from Craighead and 
others (1982), Kendall (1983), Lanner and Gilbert (these 
proceedings), Mealey (1975), Pritchard and Bobbins 
(1990), Weaver and Forcella (1986), and unpublished 
Interagency Grizzly Bear Study Team data 







Mean 


Range of 




Range 


among 


standard 


Parameters 


of means 


studies 


deviations 


Seeds per cone 


75-88 


80 


26-28 


Dry seed weight (g) 


0.07-0.10 


0.09 


0.01-0.02 


Percent protein content^ 


11.9-14.2 


12.8 




Percent ether extract^ 


21 .8-30.4 


27.1 




Percent nitrogen-free extract' 


12.4-27.2 


19.8 




Percent crude fiber' 


34.8 


(34.8) 




Percent total dietary fiber' ^ 


40.3 


(40.3) 




Percent digestible energy 


35O.I/M8.7 


49.4 


31.4 



'Based on whole seed dry weight. 
^Calculated for pinyon pine (Pinus edulis) seeds. 
^Calculated based on laboratory feeding trials (Pritchard and Robbins 
1990). 

^Calculated from differential between field samples of fresh whole material 
and feces (Mealey 1975). 



212 




Figure 1 — Distribution of whitebark pine and use of whitebark pine seeds by bears 
in North America. Sparser Canadian distribution of whitebark pine is indicated by 
dotted lines. Three different levels of pine seed use (<1 percent, 1-10 percent, and 
>10 percent frequency in bear feces) are also indicated. 



213 



with continental climates compared to areas with mari- 
time climates (2.7 percent) or areas in Canada where 
whitebark pine was sparse (2.8 percent) (df = 2, G = 987.3, 
P< 0.001). 

These differences are logically attributable to several 
factors. In areas where whitebark pine is scarce, the 
bears may not consider it to be a food, or not choose to use 
its seeds if they are disjunct from other higher value feed- 
ing opportunities. Areas influenced by maritime climates 
are characterized by an abundance of berry-producing 
shrubs, typically at elevations much lower than whitebark 
pine (Jonkel and Cowan 1971; Poelker and Hartwell 1973; 
Servheen 1983; Tisch 1961; and others). These berries 
are probably more valuable than whitebark pine seeds, 
and so local bears would logically tend to use berry-rich 
habitats to the exclusion of whitebark pine habitats (Tisch 
1961). Conversely, in areas with continental climates ber- 
ries are typically fewer and whitebark pine more abun- 
dant (Aune cuid Kasworm 1989; Mattson and others 
1991). We hypothesize that for this reason bears would 
logically focus more on the consumption of whitebark pine 
seeds. Recent heavy whitebark pine mortality in mari- 
time areas, caused by white pine blister rust, has also ar- 
guably had negative effects on bear use of pine seeds in 
these areas (Kendall and Arno 1990; Mattson and Jonkel 
1990). 

Because the large majority of whitebark pine seed use 
by bears has occurred in interior ecosystems, our subse- 
quent discussion will focus on these areas; and because 
most research concerning bear use of pine seeds comes 
from the Yellowstone ecosystem, most of the results pre- 
sented here are from this study area. 




APR MAY JUN JUL AUG SEP OCT 
n = 5 10 14 14 14 11 6 



Figure 2 — Mean percent volume of pine 
seeds in Yellowstone grizzly bear feces by 
month, 1977-90. Maximum recorded volume 
is indicated by dashed lines and mean per- 
cent volume of pine seeds in feces of occur- 
rence is shown in the inset. 



Temporal Variation in Use 

Seasonal Variation — Because whitebark pine seeds 
are durable, bears use not only seeds contemporary to 
their maturation, but also seeds that survive for up to 
1 year after initial availability of a large crop (Kendall 
1983; Mattson and Jonkel 1990; Mattson and others 1993). 
Current-year use usually begins after mid-August with 
onset of caching by red squirrels (Hutchins and Lanner 
1982), but can vary from late July to early September, 
apparently depending on seed maturation (Mattson and 
others 1993). Use of overwintered seeds does not usually 
begin imtil after mid-May (Mattson and Jonkel 1990), but 
can begin as early as the first week of April and involve 
excavation of cones out of squirrel middens, through snow 
0.3-1.2 m deep (Kendall 1983). 

Use of both current-year's and overwintered seeds by 
bears results in a primary September-October peak and 
a secondary June peak in frequency of pine seeds in bear 
feces, when averaged over a number of years (fig. 2). 
With the exception of July, bear feces tend to consist 
almost exclusively of pine seed remains (>80 percent) 
when bears are in the whitebark pine zone feeding on 
pine seeds (fig. 2). 

Annual Variation — Use of pine seeds by bears varies 
considerably among years. In the Yellowstone area, pine 
seeds have dominated the diet during some years (for ex- 
ample 1979) and been virtually absent other years (for 
example 1977, 1982, and 1988) (Mattson and others 1991, 
1992). This variation in use is related to seed crop size, 
with heavy use occurring only when crops average > 13-23 
cones per tree (from counts on permanent transects), ap- 
proaching an asymptote of use (75 percent frequency in 
feces) at crop sizes averaging 38 cones per tree. Use of 
overwintered pine seeds during the following June-July is 
similarly related to crop size, but with inflection points of 
20-29 seeds and an asymptote of 38 percent fi-equency in 
feces (fig. 3). This relationship between bear use and crop 
size is acutely sigmoidal, suggesting the existence of a use 
threshold that is possibly related to sattiration of demand 
by competing birds and rodents, nonlinear increases in 
foraging efficiency, or both. The higher thresholds and 
lower asymptote associated with use of overwintered 
seeds is attributable to depletion of the crop diiring the 
previous fall and winter. 

Varying inflection points in the hypothesized relation- 
ship between bear use and seed crop size reflect an appar- 
ent trend over the years 1980-91 toward lower thresholds 
to substantial bear use (fig. 3). The years 1980 and 1981 
are further distinguished by the relative insensitivity of 
bear use to crop size compared to later years. Interest- 
ingly, asymptotes associated with maximimi potential 
bear use have not varied significantly diiring this same 
time period. At the very least, this variance suggests 
that the relative intensity with which bears use pine 
seeds in a given study area can change. The causes of 
these apparent changes can only be speculated, including: 
increasingly efficient use of comparable-sized crops by the 
Yellowstone bear population; and an increasing willing- 
ness to use seeds imder less optimal conditions, attribut- 
able either to increasing bear densities or decreasing over- 
all carrying capacity. 



214 



AUGUST-OCTOBER USE, CONTEMPORARY YEAR 




JUNE-JULY USE, FOLLOWING YEAR 



100 



75 



50 



25 — 



-[((CONES/60)-1)/I]'^ r^ = 0 99 
% FREQ = 38.0 * e f = 353 3 



0 




B 



10 20 30 40 

MEAN NUMBER OF CONES PER TREE 



50 



Figure 3 — A, relationship between frequency 
of whitebark pine seeds in August-October 
Yellowstone grizzly bear feces and mean num- 
ber of cones counted per tree on permanent 
transects during the contemporary year, 1980- 
91 . B, relationship between frequency of pine 
seeds in June-July feces and mean number of 
cones counted the previous fall, 1 980-91 . 



Interestingly, in a part of Montana characterized by 
a more maritime climate and relatively little bear use of 
pine seeds, Tisch (1961) counted an average of five and 
16 cones per tree during 1959 and 1960, respectively. 
Both of these averages were less than the approximately 
20 cones per tree associated with the onset of heavy pine 
seed use in the Yellowstone area. 

Habitats Associated With Use 

Elevation and Aspect — As would be expected, bear 
use of whitebark pine seeds is restricted to elevations 
above the lower limits of whitebark pine distribution. 
This lower elevational limit varies with latitude and 
longitude, from about 2,400 m in the Yellowstone area 



(44.5°N.). to 1,890 m along the East Front of the Montana 
Rockies (47.5° N.) (Mattson and Jonkel 1990). 

However, within this constraint bear use varies consid- 
erably among aspects and elevations, depending on the 
year. We found that diiring three successive large seed 
crops in the Yellowstone area, bear use was nonrepetitive 
among broad aspectual and elevational categories; during 
1985-86 heaviest on east aspects at mid-elevations, during 
1987 heaviest on west and north aspects at low- to mid- 
elevations, and dxuing 1989-90 heaviest on east and south 
aspects at high elevations (Mattson and others 1993). We 
also found differences in elevational distribution of pine 
seed use between two successive crops at a local scale, 
in the Cooke City area northeast of Yellowstone National 
Park, with use concentrated at elevations >2,870 m 
dxoring 1990 and at elevations <2,720 m during 1991 
(Reinhart and Mattson 1992). This geographic variation 
was inconclusively related to seed production, but demon- 
strated the potential for considerable differences in habi- 
tat distribution of pine seed use over a 6- to 7-year period. 

Habitat Tjrpes — Ecosystems in the Western United 
States have been described in terms of ecological land- 
scape units called habitat types, that are further aggre- 
gated as series and split into phases (Pfister and others 
1977; Steele and others 1983). In the Yellowstone area 
most bear use of pine seeds has occurred with varying in- 
tensity in 14 different habitat types. In accord with varia- 
tion in distribution of pine seed use among aspects and el- 
evations, distribution of pine seed use by bears has also 
varied among habitat types (Mattson and others 1993). 
Even so, the Abies lasiocarpa /Vaccinium scoparium- 
Pinus albicaulis phase, and the Abies lasiocarpa / 
Calamagrostis canadensis and Abies lasiocarpa I 
Thalictrum occidentale habitat types have been used 
intensively during all years when pine seeds are abtm- 
dant (Mattson and Jonkel 1990). Conversely, use of habi- 
tat types at lowest and highest elevations of the white- 
bark pine zone has been most varied. Habitats in interior 
ecosystems where whitebark pine is a chmax overstory 
species have been only rarely used by bears for foraging 
on pine seeds (Mattson and Jonkel 1990). However, even 
in these habitats heavy use can occur, as diiring 1989-90 
in the Yellowstone area (Mattson and others 1993). 

Distribution of pine seed use among habitat types 
was surprisingly consistent between study areas in the 
Yellowstone ecosystem and along the East Front of the 
Montana Rocky Mountains (Mattson and Jonkel 1990). 
This suggests that patterns of habitat use associated with 
bear consumption of pine seeds are consistent among 
areas vvdth continental climates, and that more detailed 
results pertaining to habitat use from the Yellowstone 
ecosystem are extrapolable over interior regions of the 
United States. Because very Uttle information is avail- 
able concerning bear use of pine seeds in more maritime 
ecosystems, we can make no inferences about patterns of 
use in these areas other than it being restricted to sites 
with mature cone-producing whitebark pine. 

Red Squirrel and Whitebark Pine Abundance — 

Red squirrel density is probably the most important proxi- 
mal factor affecting bear use of pine seeds in interior eco- 
systems. The importance of red squirrels to especially 



215 



grizzly bear foraging has been observed by Kendall (1983), 
Mattson and Jonkel (1990), and Schallenberg and Jonkel 
(1980). Craighead and others (1982), Graber and White 
(1983), Murie (1944), and Raine and Kansas (1990) also 
mentioned rodent or squirrel caches as potential sources 
of pine seeds for bears, but were more equivocal about the 
overall importance of these intermediaries. Interestingly, 
neither Craighead and Craighead (1972) nor Mealey 
(1975) mention that grizzly bears raided red squirrel food 
caches during their studies in Yellowstone. The disparity 
between these and later observations in the Yellowstone 
area is not easily reconciled, and suggests that differences 
could be at least partly attributable to variation in bear 
behavior and pine seed availability described earlier. 

Most (86 percent) of 232 sites where grizzly bears were 
known to have used pine seeds in the Yellowstone area, 
1986-91, involved bears excavating red squirrel cone 
caches. All but five of the other 32 instances where griz- 
zly acquired pine seeds by scavenging fallen cones oc- 
curred during 1989-90 when much of the cone production 
occurred at elevations above the normal range of red 
squirrels (Mattson and others 1993; Reinhart and 
Mattson 1990). Conversely, 51 percent of 53 sites where 
black and grizzly bears used pine seeds near Cooke City, 
MT, during 1990-91 were characterized by scavenging of 
fallen cones (Reinhart and Mattson 1992). This observa- 
tion suggests that scavenging of fallen cones can be locally 
important. However, all 26 instances when bears scav- 
enged fallen cones in this area occurred during 1990, in 
conjunction with use of seeds overwintered from the im- 
usually high-elevation crop of 1989. 

Scavenging apparently predominates over excavation 
of squirrel caches in environments where squirrels are 
rare. Red squirrel densities in the whitebark pine zone 
are strongly related to a site favorability index (r^ = 0.79), 
with squirrel densities increasing as exposm*e to prevail- 
ing winds and elevation decreases (Reinhart and Mattson 
1990). Given that basal area and exclusivity of whitebark 
pine increase with wind exposure and elevation (Mattson 
and Reinhart 1990), there is an inherently negative rela- 
tionship between dominance of whitebark pine and den- 
sity of red squirrels within the whitebark pine zone. Thus, 
scavenging predominates over midden excavation where 
whitebark pine comprises >75 percent of the overstory 
(four categories of whitebark pine percentage; df = 3, 
G = 69.7, P < 0.001), and at elevations >2,870 m (four 
elevation categories; df = 3, G = 42.4, P < 0.001); where 
red squirrels are rare. 

Bear use of pine seeds is clearly the result of opportuni- 
ties implicit to the varying abundance of red squirrels and 
whitebark pine. However, there is reason to hypothesize 
that squirrel densities impose greater constraints on bear 
use of pine seeds than does abundance of whitebark pine. 
We analyzed data from 1987 to examine this hypothesis, 
looking at the relationship between unit area volume of 
excavations for pine seeds by Yellowstone grizzly bears 
and red squirrel density and whitebark pine basal area 
(fig. 4). At any given squirrel density, bear use abruptly 
increased and approached maximum levels at relatively 
low basal areas of whitebark pine, between approximately 
3 and 8 m^/ha. Conversely, the relationship between bear 
activity and red squirrel densities was more nearly linear. 



BEAR USE = 22.6 • (e 
^Opr 

< I 



-[((SD/3.68)-1)/0.382]' 



)*(e 



-[((BA/254)-1)/0.983l* 



< 
ui 

QC 
< 



20 



< 

(0 

< 
m 

UJ 

z 

a. 

bi. 
a. 
< 
m 

LU 
X 

5 



10 





























1.0 2.0 3.0 ^.0 

SQUIRREL DENSITY (N/KM) (SD) 

Figure 3 — The relationship between intensity of 
grizzly bear feeding on pine seeds, as indicated by 
density of excavations (m^/km) and squirrel density 
and whitebark pine basal area, Yellowstone area, 
1987. 



with excavations gradually increasing to a maximum that 
corresponded to maximum squirrel densities observed in 
the whitebark pine zone. Basically, whitebark pine basal 
area seemed to be associated primarily with the probabil- 
ity that a given midden would be excavated, sometimes 
repeatedly, by bears (Mattson and Reinhart 1987), with 
the greatest effects at very low whitebark pine abun- 
dance. Once this low threshold of whitebark pine abun- 
dance was exceeded (> about 4.3 m^/ha), red squirrel den- 
sities appeared to limit bear use of pine seeds during most 
years. 

Timber Stand Age — Bear use of whitebark pine seeds 
in the Yellowstone area has occurred almost exclusively in 
mature stands. According to a successional classification 
of timber types by Despain (1990), all bear use of pine 
seeds from squirrel middens and most bear use of scav- 
enged seeds has occurred in "mature" cover types, typi- 
cally designated LP2, LP3, SF, WB2, WB3, WB, or DF3. 
Results from a use-availability analysis of cover types in 
Yellowstone National Park corroborate the affinity of griz- 
zly bears for later successional cover types when using 
pine seeds (Knight and others 1984). This analysis showed 
that the early successional WBl cover type was consis- 
tently imder-used or used without preference during all 
seasons, whereas mature WB cover types were consistently 
used with disproportionally greater intensity than ex- 
pected, especially dining the fall, primarily for the con- 
sumption of pine seeds. 



216 



Based on ages given by Despain (1990) cover types asso- 
ciated with pine seed use are rarely less than 100 years 
old. Ages of 160 stands in two intensive Yellowstone 
study areas located in the whitebark pine zone (Mattson 
and Reinhart 1987) also revealed that none of these ma- 
ture cover types was <125 years old. 

These minimum stand ages correspond with use ex- 
pected by levels of cone production. Weaver and others 
(1990) observed that whitebark pine cone production was 
relatively low in stands <100 years old and reached peak 
levels soon thereafter. Morgan and Biinting (in press) 
also observed that younger WBO and WBl stands pro- 
duced few cones and had few excellent crops. Both of 
these authors observed that cone production persisted at 
least at moderate levels in much older stands, up to 300 to 
400 years old. These observations correspond well with 
the resvilts of more exhaustive studies of the closely re- 
lated Siberian stone pine (Pinus sibirica) (Axelrod 1986) 
by Russian researchers (Mattson and Jonkel 1990). 

These relationships of use to stand age have several im- 
plications. First, potential productivity of whitebark pine 
sites, at least for bears, is not realized until after approxi- 
mately 100 years, and can persist for 200 to 300 years 
thereafter. Thus, in contrast to many other temperate 
ecosystems where production of bear foods is concentrated 
on berry-producing shrubs in early successional stages, 
production of bear foods in the whitebark pine zone is con- 
centrated in much older mid- to late-successional stands 
(Mattson 1990). Second, given these time frames, long- 
term productivity of the whitebark pine zone in interior 
ecosystems would be best assiired by disturbance regimes 
that replaced approximately 3 percent of mature acreage 
per decade. 

EFFECTS ON BEAR DEMOGRAPHY 
Fecundity 

Whitebark pine seeds are an important so\irce of energy 
for bears that live in interior ecosystems of the United 
States. Craighead and others (1982) estimated that pine 
seeds were the most important soxirce of energy for grizzly 
bears in their study area located east of the Continental 
Divide in the Rocky Mountains of Montana. Preliminary 
estimates by Servheen and others (1986) suggested that 
Yellowstone grizzly bears derived about 23 percent of 
their net digested energy from whitebark pine seeds, 
second only to the approximately 30 percent